CySecurity News - Latest Information Security and Hacking Incidents: Google+

Artifcial Intelligence CyberCrime Cybersecurity CyberThreat Google Google Gemini AI LLM Phishing Content Technology

Google Gemini Exploit Enables Covert Delivery of Phishing Content

An AI-powered automation system in professional environments, such as Google Gemini for Workspace, is vulnerable to a new security flaw. Using Google’s advanced large language model (LLM) integration within its ecosystem, Gemini enables the use of artificial intelligence (AI) directly with a wide range of user tools, including Gmail, to simplify workplace tasks.

A key feature of the app is the ability to request concise summaries of emails, which are intended to save users time and prevent them from becoming fatigued in their inboxes by reducing the amount of time they spend in it. Security researchers have, however identified a significant flaw in this feature which appears to be so helpful.

As Mozilla bug bounty experts pointed out, malicious actors can take advantage of the trust users place in Gemini's automated responses by manipulating email content so that the AI is misled into creating misleading summaries by manipulating the content. As a result of the fact that Gemini operates within Google's trusted environment, users are likely to accept its interpretations without question, giving hackers a prime opportunity. This finding highlights what is becoming increasingly apparent in the cybersecurity landscape: when powerful artificial intelligence tools are embedded within widely used platforms, even minor vulnerabilities can be exploited by sophisticated social engineers.

It is the vulnerability at the root of this problem that Gemini can generate e-mail summaries that seem legitimate but can be manipulated so as to include deceptive or malicious content without having to rely on conventional red flags, such as suspicious links or file attachments, to detect it.

An attack can be embedded within an email body as an indirect prompt injection by attackers, according to cybersecurity researchers. When Gemini's language model interprets these hidden instructions during thesummarisationn process, it causes the AI to unintentionally include misleading messages in the summary that it delivers to the user, unknowingly.

As an example, a summary can falsely inform the recipient that there has been a problem with their account, advising them to act right away, and subtly direct them to a phishing site that appears to be reliable and trustworthy.

While prompt injection attacks on LLMs have been documented since the year 2024, and despite the implementation of numerous safeguards by developers to prevent these manipulations from occurring, this method continues to be effective even today. This tactic is persisting because of the growing sophistication of threat actors as well as the challenge of fully securing generative artificial intelligence systems that are embedded in critical communication platforms.

There is also a need to be more vigilant when developing artificial intelligence and making sure users are aware of it, as traditional cybersecurity cues may no longer apply to these AI-driven environments. In order to find these vulnerabilities, a cybersecurity researcher, Marco Figueroa, identified them and responsibly disclosed them through Mozilla's 0Din bug bounty program, which specialises in finding vulnerabilities in generative artificial intelligence.

There is a clever but deeply concerning method of exploitation demonstrated in Figueroa's proof-of-concept. The attack begins with a seemingly harmless e-mail sent to the intended victim that appears harmless at first glance. A phishing prompt disguised in white font on a white background is hidden in a secondary, malicious component of the message, which conceals benign information so as to avoid suspicion of the message.

When viewed in a standard email client, it is completely invisible to the human eye and is hidden behind benign content. The malicious message is strategically embedded within custom tags, which are not standard HTML elements, but which appear to be interpreted in a privileged manner by Gemini's summarization function, as they are not standard HTML elements.

By activating the "Summarise this email" feature in Google Gemini, a machine learning algorithm takes into account both visible and hidden text within the email. Due to the way Gemini handles input wrapped in tags, it prioritises and reproduces the hidden message verbatim within the summary, placing it at the end of the response, as it should.

In consequence, what appears to be a trustworthy, AI-generated summary now contains manipulative instructions which can be used to entice people to click on phishing websites, effectively bypassing traditional security measures. A demonstration of the ease with which generative AI tools can be exploited when trust in the system is assumed is demonstrated in this attack method, and it further demonstrates the importance of robust sanitisation protocols as well as input validation protocols for prompt sanitisation.

It is alarming how effectively the exploitation technique is despite its technical simplicity. An invisible formatting technique enables the embedding of hidden prompts into an email, leveraging Google Gemini's interpretation of raw content to capitalise on its ability to comprehend the content. In the documented attack, a malicious actor inserts a command inside a span element with font-size: 0 and colour: white, effectively rendering the content invisible to the recipient who is viewing the message in a standard email client.

Unlike a browser, which renders only what can be seen by the user, Gemini process the entire raw HTML document, including all hidden elements. As a consequence, Gemini's summary feature, which is available to the user when they invoke it, interprets and includes the hidden instruction as though it were part of the legitimate message in the generated summary.

It is important to note that this flaw has significant implications for services that operate at scale, as well as for those who use them regularly. A summary tool that is capable of analysing HTML inline styles, such as font-size:0, colour: white, and opacity:0, should be instructed to ignore or neutralise these styles, which render text visually hidden.

The development team can also integrate guard prompts into LLM behaviour, instructing models not to ignore invisible content, for example. In terms of user education, he recommends that organisations make sure their employees are aware that AI-generated summaries, including those generated by Gemini, serve only as informational aids and should not be treated as authoritative sources when it comes to urgent or security-related instructions.

A vulnerability of this magnitude has been discovered at a crucial time, as more and more tech companies are increasingly integrating LLMs into their platforms to automate productivity. In contrast to previous models, where users would manually trigger AI tools, the new paradigm is a shift to automated AI tools that will run in the background instead.

It is for this reason that Google introduced the Gemini side panel last year in Gmail, Docs, Sheets, and other Workspace apps to help users summarise and create content within their workflow seamlessly. A noteworthy change in Gmail's functionality is that on May 29, Google enabled automatic email summarisation for users whose organisations have enabled smart features across Gmail, Chat, Meet, and other Workspace tools by activating a default personalisation setting.

As generative artificial intelligence becomes increasingly integrated into everyday communication systems, robust security protocols will become increasingly important as this move enhances convenience. This vulnerability exposes an issue of fundamental inadequacy in the current guardrails used for LLM, primarily focusing on filtering or flagging content that is visible to the user.

A significant number of AI models, including the Google Gemini AI model, continue to use raw HTML markup, making them susceptible to obfuscation techniques such as zero-font text and white-on-white formatting. Despite being invisible to users, these techniques are still considered valid input to the model by the model-thereby creating a blind spot for attackers that can easily be exploited by attackers.

Mozilla's 0Din program classified the issue as a moderately serious vulnerability by Mozilla, and said that the flaw could be exploited by hackers to harvest credential information, use vishing (voice-phishing), and perform other social engineering attacks by exploiting trust in artificial intelligence-generated content in order to gain access to information.

In addition to the output filter, a post-processing filter can also function as an additional safeguard by inspecting artificial intelligence-generated summaries for signs of manipulation, such as embedded URLs, telephone numbers, or language that implies urgency, flagging these suspicious summaries for human review. This layered defence strategy is especially vital in environments where AI operates at scale.

As well as protecting against individual attacks, there is also a broader supply chain risk to consider. It is clear that mass communication systems, such as CRM platforms, newsletters, and automated support ticketing services, are potential vectors for injection, according to researcher Marco Figueroa. There is a possibility that a single compromised account on any of these SaaS systems can be used to spread hidden prompt injections across thousands of recipients, turning otherwise legitimate SaaS services into large-scale phishing attacks.

There is an apt term to describe "prompt injections", which have become the new email macros according to the research. The exploit exhibited by Phishing for Gemini significantly underscores a fundamental truth: even apparently minor, invisible code can be weaponised and used for malicious purposes.

As long as language models don't contain robust context isolation that ensures third-party content is sandboxed or subjected to appropriate scrutiny, each piece of input should be viewed as potentially executable code, regardless of whether it is encoded correctly or not. In light of this, security teams should start to understand that AI systems are no longer just productivity tools, but rather components of a threat surface that need to be actively monitored, measured, and contained.

The risk landscape of today does not allow organisations to blindly trust AI output. Because generative artificial intelligence is being integrated into enterprise ecosystems in ever greater numbers, organisations must reevaluate their security frameworks in order to address the emerging risks that arise from machine learning systems in the future.

Considering the findings regarding Google Gemini, it is urgent to consider AI-generated outputs as potential threat vectors, as they are capable of being manipulated in subtle but impactful ways. A security protocol based on AI needs to be implemented by enterprises to prevent such exploitations from occurring, robust validation mechanisms for automated content need to be established, and a collaborative oversight system between development, IT, and security teams must be established to ensure this doesn't happen again.

Moreover, it is imperative that AI-driven tools, especially those embedded within communication workflows, be made accessible to end users so that they can understand their capabilities and limitations. In light of the increasing ease and pervasiveness of automation in digital operations, it will become increasingly essential to maintain a culture of informed vigilance across all layers of the organisation to maintain trust and integrity.

Google Gemini Bug Exploits Summaries for Phishing Scams



 

Technology

AI AI Prompt ChatGPT Gemini Google Internet Technology

Google Gemini Bug Exploits Summaries for Phishing Scams

False AI summaries leading to phishing attacks

Google Gemini for Workspace can be exploited to generate email summaries that appear legitimate but include malicious instructions or warnings that direct users to phishing sites without using attachments or direct links.

Google Gemini for Workplace can be compromised to create email summaries that look real but contain harmful instructions or warnings that redirect users to phishing websites without using direct links or attachments.

Similar attacks were reported in 2024 and afterwards; safeguards were pushed to stop misleading responses. However, the tactic remains a problem for security experts.

Gemini for attack

A prompt-injection attack on the Gemini model was revealed via cybersecurity researcher Marco Figueoa, at 0din, Mozilla’s bug bounty program for GenAI tools. The tactic creates an email with a hidden directive for Gemini. The threat actor can hide malicious commands in the message body text at the end via CSS and HTML, which changes the font size to zero and color to white.

According to Marco, who is GenAI Bug Bounty Programs Manager at Mozilla, “Because the injected text is rendered in white-on-white (or otherwise hidden), the victim never sees the instruction in the original message, only the fabricated 'security alert' in the AI-generated summary. Similar indirect prompt attacks on Gemini were first reported in 2024, and Google has already published mitigations, but the technique remains viable today.”

Gmail does not render the malicious instruction as there are no attachments or links present, and the message may reach the victim’s inbox. If the receiver opens the email and asks Gemini to make a summary of the received mail, the AI tool will parse the invisible directive and create the summary. Figueroa provides an example of Gemini following hidden prompts, accompanied by a security warning that the victim’s Gmail password and phone number may be compromised.

Impact

Supply-chain threats: CRM systems, automated ticketing emails, and newsletters can become injection vectors, changing one exploited SaaS account into hundreds of thousands of phishing beacons.

Cross-product surface: The same tactics applies to Gemini in Slides, Drive search, Docs and any workplace where the model is getting third-party content.

According to Marco, “Security teams must treat AI assistants as part of the attack surface and instrument them, sandbox them, and never assume their output is benign.”

Security Breach Reveals "Catwatchful" Spyware is Snooping on Users



 

Technology

AI Cloud Data Google Internet Security Breach Technology

Security Breach Reveals "Catwatchful" Spyware is Snooping on Users

A security bug in a stealthy Android spyware operation, “Catwatchful,” has exposed full user databases affecting its 62,000 customers and also its app admin. The vulnerability was found by cybersecurity expert Eric Daigle reported about the spyware app’s full database of email IDs and plaintext passwords used by Catwatchful customers to access stolen data from the devices of their victims.

Most of the victims were based in India, Argentina, Peru, Mexico, Colombia, Bolivia, and Ecuador. A few records date back to 2018. The leaked database also revealed the identity of the Catwatchful admin called Omar Soca Char.

The Catwatchful database also revealed the identity of the spyware operation’s administrator, Omar Soca Charcov, a developer based in Uruguay.

About Catwatchful

Catwatchful is a spyware that pretends to be a child monitoring app, claiming to be “invisible and can not be detected,” while it uploads the victim’s data to a dashboard accessible to the person who planted the app. The stolen data includes real-time location data, victims’ photos, and messages. The app can also track live ambient audio from the device’s mic and access the phone camera (both front and rear).

Catwatchful and similar apps are banned on app stores, and depend on being downloaded and deployed by someone having physical access to a victim’s phone. These apps are famous as “stalkerware” or “spouseware” as they are capable of unauthorized and illegal non-consensual surveillance of romantic partners and spouses.

Rise of spyware apps

The Catwatchful incident is the fifth and latest in this year’s growing list of stalkerware scams that have been breached, hacked, or had their data exposed.

How was the spyware found?

Daigle has previously discovered stalkerware exploits. Catwatchful uses a custom-made API, which the planted app uses to communicate to send data back to Catwatchful servers. The stalkerware also uses Google Firebase to host and store stolen data.

According to Techradar, the “data was stored on Google Firebase, sent via a custom API that was unauthenticated, resulting in open access to user and victim data. The report also confirms that, although hosting had initially been suspended by HostGator, it had been restored via another temporary domain."

Ditch Passwords, Use Passkeys to Secure Your Account



 

Windows

Apple Google Internet Microsoft Passkeys Passwords Technology Windows

Ditch Passwords, Use Passkeys to Secure Your Account

Ditch passwords, use passkeys

Microsoft and Google users, in particular, have been warned about ditching passwords for passkeys. Passwords are easy to steal and can unlock your digital life. Microsoft has been at the forefront, confirming it will delete passwords for more than a billion users. Google, too, has warned that most of its users will have to add passkeys to their accounts.

What are passkeys?

Instead of a username and password, passkeys use our device security to log into our account. This means that there is no password to hack and no two-factor authentication codes to bypass, making it phishing-resistant.

At the same time, the Okta team warned that it found threat actors exploiting v0, an advanced GenAI tool made by Vercelopens, to create phishing websites that mimic real sign-in webpages

Okta warns users to not use passwords

A video shows how this works, raising concerns about users still using passwords to sign into their accounts, even when backed by multi-factor authentication, and “especially if that 2FA is nothing better than SMS, which is now little better than nothing at all,” according to Forbes.

According to Okta, “This signals a new evolution in the weaponization of GenAI by threat actors who have demonstrated an ability to generate a functional phishing site from simple text prompts. The technology is being used to build replicas of the legitimate sign-in pages of multiple brands, including an Okta customer.”

Why are passwords not safe?

It is shocking how easy a login webpage can be mimicked. Users should not be surprised that today’s cyber criminals are exploiting and weaponizing GenAI features to advance and streamline their phishing attacks. AI in the wrong hands can have massive repercussions for the cybersecurity industry.

According to Forbes, “Gone are the days of clumsy imagery and texts and fake sign-in pages that can be detected in an instant. These latest attacks need a technical solution.”

Users are advised to add passkeys to their accounts if available and stop using passwords when signing in to their accounts. Users should also ensure that if they use passwords, they should be long and unique, and not backed up by SMS 2-factor authentication.

New Android Feature Detects Fake Mobile Networks



 

stingrays

Android App privacy Google Mobile Security stingrays

New Android Feature Detects Fake Mobile Networks

In a critical move for mobile security, Google is preparing to roll out a new feature in Android 16 that will help protect users from fake mobile towers, also known as cell site simulators, that can be used to spy on people without their knowledge.

These deceptive towers, often referred to as stingrays or IMSI catchers, are devices that imitate real cell towers. When a smartphone connects to them, attackers can track the user’s location or intercept sensitive data like phone calls, text messages, or even the phone's unique ID numbers (such as IMEI). What makes them dangerous is that users typically have no idea their phones are connected to a fraudulent network.

Stingrays usually exploit older 2G networks, which lack strong encryption and tower authentication. Even if a person uses a modern 4G or 5G connection, their device can still switch to 2G if the signal is stronger opening the door for such attacks.

Until now, Android users had very limited options to guard against these silent threats. The most effective method was to manually turn off 2G network support—something many people aren’t aware of or don’t know how to do.

That’s changing with Android 16. According to public documentation on the Android Open Source Project, the operating system will introduce a “network security warning” feature. When activated, it will notify users if their phone connects to a mobile network that behaves suspiciously, such as trying to extract device identifiers or downgrade the connection to an unsecured one.

This feature will be accessible through the “Mobile Network Security” settings, where users can also manage 2G-related protections. However, there's a catch: most current Android phones, including Google's own Pixel models, don’t yet have the hardware required to support this function. As a result, the feature is not yet visible in settings, and it’s expected to debut on newer devices launching later this year.

Industry observers believe that this detection system might first appear on the upcoming Pixel 10, potentially making it one of the most security-focused smartphones to date.

While stingray technology is sometimes used by law enforcement agencies for surveillance under strict regulations, its misuse remains a serious privacy concern especially if such tools fall into the wrong hands.

With Android 16, Google is taking a step toward giving users more control and awareness over the security of their mobile connections. As surveillance tactics become more advanced, these kinds of features are increasingly necessary to protect personal privacy.

WhatsApp Under Fire for AI Update Disrupting Group Communication



 

Privacy

AI AI technology ChatGPT Google Hacking WhatsApp Messanger Meta AI Privacy

WhatsApp Under Fire for AI Update Disrupting Group Communication

The new artificial intelligence capability introduced by WhatsApp aims to transform the way users interact with their conversations through sophisticated artificial intelligence. It uses advanced technology from Meta AI to provide a concise summary of unread messages across individual chats as well as group chats, which is referred to as Message Summaries.

The tool was created to help users stay informed in increasingly active chat environments by automatically compiling key points and contextual highlights, allowing them to catch up in just a few clicks without having to scroll through lengthy message histories to catch up. The company claims all summaries are generated privately, so that confidentiality can be maintained and the process of use is as simple as possible for the user.

WhatsApp announces its intention of integrating artificial intelligence-driven solutions into its app to improve user convenience as well as reshape communication habits for its global community with this rollout, sparking both excitement and controversy as a result. Despite being announced last month, WhatsApp’s innovative Message Summaries feature has moved from pilot testing to a full-scale rollout after successfully passing pilot testing.

Having refined the tool and collected feedback from its users, it is now considered to be stable and has been formally launched for wider use. In the initial phase, the feature is only available to US users and is restricted to the English language at this time. This indicates that WhatsApp is cautious when it comes to deploying large-scale artificial intelligence.

Nevertheless, the platform announced plans to extend its availability to more regions at some point in the future, along with the addition of multilingual support. The phased rollout strategy emphasises that the company is focused on ensuring that the technology is reliable and user-friendly before it is extended to the vast global market.

It is WhatsApp's intention to focus on a controlled release so as to gather more insights about users' interaction with the AI-generated conversation summaries, as well as to fine-tune the experience before expanding internationally. As a result of WhatsApp's inability to provide an option for enabling or concealing the Message Summaries feature, there has been a significant amount of discontent among users.

Despite the fact that Meta has refused to clarify the reason regarding the lack of an opt-out mechanism or why users were not offered the opportunity to opt out of the AI integration, they have not provided any explanation so far. As concerning as the technology itself is, the lack of transparency has been regarded equally as a cause for concern by many, raising questions about the control people have over their personal communications. As a result of these limitations, some people have attempted to circumvent the chatbot by switching to a WhatsApp Business account as a response.

In addition, several users have commented that this strategy removed the AI functionality from Meta AI, but others have noted that the characteristic blue circle, which indicates Meta AI's presence, still appeared, which exacerbated the dissatisfaction and uncertainty.

The Meta team hasn’t confirmed whether the business-oriented version of WhatsApp will continue to be exempt from AI integration for years to come. This rollout also represents Meta’s broader goal of integrating generative AI into all its platforms, which include Facebook and Instagram, into its ecosystem.

Towards the end of 2024, Meta AI was introduced for the first time in Facebook Messenger in the United Kingdom, followed by a gradual extension into WhatsApp as part of a unified vision to revolutionise digital interactions. However, many users have expressed their frustration with this feature because it often feels intrusive and ultimately is useless, despite these ambitions.

The chatbot appears to activate frequently when individuals are simply searching for past conversations or locating contacts, which results in obstructions rather than streamlining the experience. According to the initial feedback received, AI-generated responses are frequently perceived as superficial, repetitive, or even irrelevant to the conversation's context, as well as generating a wide range of perceptions of their value.

A Meta AI platform has been integrated directly into WhatsApp, unlike standalone platforms such as ChatGPT and Google Gemini, which are separately accessible by users. WhatsApp is a communication application that is used on a daily basis to communicate both personally and professionally. Because the feature was integrated without explicit consent and there were doubts about its usefulness, many users are beginning to wonder whether such pervasive AI assistance is really necessary or desirable.

It has also been noted that there is a growing chorus of criticism about the inherent limitations of artificial intelligence in terms of reliably interpreting human communication. Many users have expressed their scepticism about AI's ability to accurately condense even one message within an active group chat, let alone synthesise hundreds of exchanges. It is not the first time Apple has faced similar challenges; Apple has faced similar challenges in the past when it had to pull an AI-powered feature that produced unintended and sometimes inaccurate summaries.

As of today, the problem of "hallucinations," which occur in the form of factually incorrect or contextually irrelevant content generated by artificial intelligence, remains a persistent problem across nearly every generative platform, including commonly used platforms like ChatGPT. Aside from that, artificial intelligence continues to struggle with subtleties such as humour, sarcasm, and cultural nuance-aspects of natural conversation that are central to establishing a connection.

In situations where the AI is not trained to recognise offhand or joking remarks, it can easily misinterpret those remarks. This leads to summaries that are alarmist, distorted, or completely inaccurate, as compared to human recipients' own. Due to the increased risk of misrepresentation, users who rely on WhatsApp for authentic, nuanced communication with colleagues, friends, and family are becoming more apprehensive than before.

A philosophical objection has been raised beyond technical limitations, stating that the act of participating in a conversation is diminished by substituting real engagement for machine-generated recaps. There is a shared sentiment that the purpose of group chats lies precisely in the experience of reading and responding to the genuine voices of others while scrolling through a backlog of messages.

However, there is a consensus that it is exhausting to scroll through such a large backlog of messages. It is believed that the introduction of Message Summaries not only threatens clear communication but also undermines the sense of personal connection that draws people into these digital communities in the first place, which is why these critics are concerned.

In order to ensure user privacy, WhatsApp has created the Message Summaries feature using a new framework known as Private Processing, which is designed to safeguard user privacy. Meta and WhatsApp are specifically ensuring that neither the contents of their conversations nor the summaries that the AI system produces are able to be accessed by them, which is why this approach was developed.

Instead of sending summaries to external servers, the platform is able to generate them locally on the users' devices, reinforcing its commitment to privacy. Each summary, presented in a clear bullet point format, is clearly labelled as "visible only to you," emphasising WhatsApp's privacy-centric design philosophy behind the feature as well.

Message Summaries have shown to be especially useful in group chats in which the amount of unread messages is often overwhelming, as a result of the large volume of unread messages. With this tool, users are able to remain informed without having to read every single message, because lengthy exchanges are distilled into concise snapshots that enable them to stay updated without having to scroll through each and every individual message.

The feature is disabled by default and needs to be activated manually, which addresses privacy concerns. Upon activating the feature, eligible chats display a discreet icon, signalling the availability of a summary without announcing it to other participants. Meta’s confidential computing infrastructure is at the core of its system, and in principle, it is comparable to Apple’s private cloud computing architecture.

A Trusted Execution Environment (TEE) provides a foundation for Private Processing, ensuring that confidential information is handled in an effective manner, with robust measures against tampering, and clear mechanisms for ensuring transparency are in place.

A system's architecture is designed to shut down automatically or to generate verifiable evidence of the intrusion whenever any attempt is made to compromise the security assurances of the system. As well as supporting independent third-party audits, Meta has intentionally designed the framework in such a way that it will remain stateless, forward secure, and immune to targeted attacks so that Meta's claims about data protection can be verified.

Furthermore, advanced chat privacy settings are included as a complement to these technical safeguards, as they allow users to select the conversations that will be eligible for AI-generated summaries and thus offer granular control over the use of the feature. Moreover, when a user decides to enable summaries in a chat, no notification is sent to other participants, allowing for greater discretion on the part of other participants.

There is currently a phase in which Message Summaries are being gradually introduced to users in the United States. They can only be read in English at the moment. There has been confirmation by Meta that the feature will be expanded to additional regions and supported in additional languages shortly, as part of their broader effort to integrate artificial intelligence into all aspects of their service offerings.

As WhatsApp intensifies its efforts to embed AI capabilities deeper and deeper into everyday communication, Message Summaries marks a pivotal moment in the evolution of relationships between technology and human interaction as the company accelerates its ambition to involve AI capabilities across the entire enterprise.

Even though the company has repeatedly reiterated that it is committed to privacy, transparency, and user autonomy, the response to this feature has been polarised, which highlights the challenges associated with incorporating artificial intelligence in spaces where trust, nuance, and human connection are paramount.

It is a timely reminder that, for both individuals and organisations, the growth of convenience-driven automation impacts the genuine social fabric that is a hallmark of digital communities and requires a careful assessment.

As platforms evolve, stakeholders would do well to remain vigilant with the changes to platform policies, evaluate whether such tools align with the communication values they hold dear, and consider offering structured feedback in order for these technologies to mature with maturity. As artificial intelligence continues to redefine the contours of messaging, users will need to be open to innovation while also expressing critical thought about the long-term implications on privacy, comprehension, and even the very nature of meaningful dialogue as AI use continues to grow in popularity.

OpenAI Rolls Out Premium Data Connections for ChatGPT Users



 

Artificial Intelligence ChatGPT Connectors CyberThreat Data Connections Drive Dropbox Emails Gmail Google GPT-4 OpenAI Privacy Twitter X

OpenAI Rolls Out Premium Data Connections for ChatGPT Users

The ChatGPT solution has become a transformative artificial intelligence solution widely adopted by individuals and businesses alike seeking to improve their operations. Developed by OpenAI, this sophisticated artificial intelligence platform has been proven to be very effective in assisting users with drafting compelling emails, developing creative content, or conducting complex data analysis by streamlining a wide range of workflows.

OpenAI is continuously enhancing ChatGPT's capabilities through new integrations and advanced features that make it easier to integrate into the daily workflows of an organisation; however, an understanding of the platform's pricing models is vital for any organisation that aims to use it efficiently on a day-to-day basis. A business or an entrepreneur in the United Kingdom that is considering ChatGPT's subscription options may find that managing international payments can be an additional challenge, especially when the exchange rate fluctuates or conversion fees are hidden.

In this context, the Wise Business multi-currency credit card offers a practical solution for maintaining financial control as well as maintaining cost transparency. This payment tool, which provides companies with the ability to hold and spend in more than 40 currencies, enables them to settle subscription payments without incurring excessive currency conversion charges, which makes it easier for them to manage budgets as well as adopt cutting-edge technology.

A suite of premium features has been recently introduced by OpenAI that aims to enhance the ChatGPT experience for subscribers by enhancing its premium features. There is now an option available to paid users to use advanced reasoning models that include O1 and O3, which allow users to make more sophisticated analytical and problem-solving decisions.

The subscription comes with more than just enhanced reasoning; it also includes an upgraded voice mode that makes conversational interactions more natural, as well as improved memory capabilities that allow the AI to retain context over the course of a long period of time. It has also been enhanced with the addition of a powerful coding assistant designed to help developers automate workflows and speed up the software development process.

To expand the creative possibilities even further, OpenAI has adjusted token limits, which allow for greater amounts of input and output text and allow users to generate more images without interruption. In addition to expedited image generation via a priority queue, subscribers have the option of achieving faster turnaround times during high-demand periods.

In addition to maintaining full access to the latest models, paid accounts are also provided with consistent performance, as they are not forced to switch to less advanced models when server capacity gets strained-a limitation that free users may still have to deal with. While OpenAI has put in a lot of effort into enriching the paid version of the platform, the free users have not been left out. GPT-4o has effectively replaced the older GPT-4 model, allowing complimentary accounts to take advantage of more capable technology without having to fall back to a fallback downgrade.

In addition to basic imaging tools, free users will also receive the same priority in generation queues as paid users, although they will also have access to basic imaging tools. With its dedication to making AI broadly accessible, OpenAI has made additional features such as ChatGPT Search, integrated shopping assistance, and limited memory available free of charge, reflecting its commitment to making AI accessible to the public.

ChatGPT's free version continues to be a compelling option for people who utilise the software only sporadically-perhaps to write occasional emails, research occasionally, and create simple images. In addition, individuals or organisations who frequently run into usage limits, such as waiting for long periods of time for token resettings, may find that upgrading to a paid plan is an extremely beneficial decision, as it unlocks uninterrupted access as well as advanced capabilities.

In order to transform ChatGPT into a more versatile and deeply integrated virtual assistant, OpenAI has introduced a new feature, called Connectors, which is designed to transform the platform into an even more seamless virtual assistant. It has been enabled by this new feature for ChatGPT to seamlessly interface with a variety of external applications and data sources, allowing the AI to retrieve and synthesise information from external sources in real time while responding to user queries.

With the introduction of Connectors, the company is moving forward towards providing a more personal and contextually relevant experience for our users. In the case of an upcoming family vacation, for example, ChatGPT can be instructed by users to scan their Gmail accounts in order to compile all correspondence regarding the trip. This allows users to streamline travel plans rather than having to go through emails manually.

With its level of integration, Gemini is similar to its rivals, which enjoy advantages from Google's ownership of a variety of popular services such as Gmail and Calendar. As a result of Connectors, individuals and businesses will be able to redefine how they engage with AI tools in a new way. OpenAI intends to create a comprehensive digital assistant by giving ChatGPT secure access to personal or organisational data that is residing across multiple services, by creating an integrated digital assistant that anticipates needs, surfaces critical insights, streamlines decision-making processes, and provides insights.

There is an increased demand for highly customised and intelligent assistance, which is why other AI developers are likely to pursue similar integrations to remain competitive. The strategy behind Connectors is ultimately to position ChatGPT as a central hub for productivity — an artificial intelligence that is capable of understanding, organising, and acting upon every aspect of a user’s digital life.

In spite of the convenience and efficiency associated with this approach, it also illustrates the need to ensure that personal information remains protected while providing robust data security and transparency in order for users to take advantage of these powerful integrations as they become mainstream. In its official X (formerly Twitter) account, OpenAI has recently announced the availability of Connectors that can integrate with Google Drive, Dropbox, SharePoint, and Box as part of ChatGPT outside of the Deep Research environment.

As part of this expansion, users will be able to link their cloud storage accounts directly to ChatGPT, enabling the AI to retrieve and process their personal and professional data, enabling it to create responses on their own. As stated by OpenAI in their announcement, this functionality is "perfect for adding your own context to your ChatGPT during your daily work," highlighting the company's ambition of making ChatGPT more intelligent and contextually aware.

It is important to note, however, that access to these newly released Connectors is confined to specific subscriptions and geographical restrictions. A ChatGPT Pro subscription, which costs $200 per month, is exclusive to ChatGPT Pro subscribers only and is currently available worldwide, except for the European Economic Area (EEA), Switzerland and the United Kingdom. Consequently, users whose plans are lower-tier, such as ChatGPT Plus subscribers paying $20 per month, or who live in Europe, cannot use these integrations at this time.

Typically, the staggered rollout of new technologies is a reflection of broader challenges associated with regulatory compliance within the EU, where stricter data protection regulations as well as artificial intelligence governance frameworks often delay their availability. Deep Research remains relatively limited in terms of the Connectors available outside the company. However, Deep Research provides the same extensive integration support as Deep Research does.

In the ChatGPT Plus and Pro packages, users leveraging Deep Research capabilities can access a much broader array of integrations — for example, Outlook, Teams, Gmail, Google Drive, and Linear — but there are some restrictions on regions as well. Additionally, organisations with Team plans, Enterprise plans, or Educational plans have access to additional Deep Research features, including SharePoint, Dropbox, and Box, which are available to them as part of their Deep Research features.

Additionally, OpenAI is now offering the Model Context Protocol (MCP), a framework which allows workspace administrators to create customised Connectors based on their needs. By integrating ChatGPT with proprietary data systems, organizations can create secure, tailored integrations, enabling highly specialized use cases for internal workflows and knowledge management that are highly specialized.

With the increasing adoption of artificial intelligence solutions by companies, it is anticipated that the catalogue of Connectors will rapidly expand, offering users the option of incorporating external data sources into their conversations. The dynamic nature of this market underscores that technology giants like Google have the advantage over their competitors, as their AI assistants, such as Gemini, can be seamlessly integrated throughout all of their services, including the search engine.

The OpenAI strategy, on the other hand, relies heavily on building a network of third-party integrations to create a similar assistant experience for its users. It is now generally possible to access the new Connectors in the ChatGPT interface, although users will have to refresh their browsers or update the app in order to activate the new features.

As AI-powered productivity tools continue to become more widely adopted, the continued growth and refinement of these integrations will likely play a central role in defining the future of AI-powered productivity tools. A strategic approach is recommended for organisations and professionals evaluating ChatGPT as generative AI capabilities continue to mature, as it will help them weigh the advantages and drawbacks of deeper integration against operational needs, budget limitations, and regulatory considerations that will likely affect their decisions.

As a result of the introduction of Connectors and the advanced subscription tiers, people are clearly on a trajectory toward more personalised and dynamic AI assistance, which is able to ingest and contextualise diverse data sources. As a result of this evolution, it is also becoming increasingly important to establish strong frameworks for data governance, to establish clear controls for access to the data, and to ensure adherence to privacy regulations.

If companies intend to stay competitive in an increasingly automated landscape by investing early in these capabilities, they can be in a better position to utilise the potential of AI and set clear policies that balance innovation with accountability by leveraging the efficiencies of AI in the process. In the future, the organisations that are actively developing internal expertise, testing carefully selected integrations, and cultivating a culture of responsible AI usage will be the most prepared to fully realise the potential of artificial intelligence and to maintain a competitive edge for years to come.

Russian Threat Actors Circumvent Gmail Security with App Password Theft



 

SVR

APT29 CyberCrime Cybersecurity CyberThreat Email Hacking Gmail Google GTIG MFA Password Privacy SVR

Russian Threat Actors Circumvent Gmail Security with App Password Theft

As part of Google's Threat Intelligence Group (GTIG), security researchers discovered a highly sophisticated cyber-espionage campaign orchestrated by Russian threat actors. They succeeded in circumventing Google's multi-factor authentication (MFA) protections for Gmail accounts by successfully circumventing it.

A group of researchers found that the attackers used highly targeted and convincing social engineering tactics by impersonating Department of State officials in order to establish trust with their victims in the process. As soon as a rapport had been built, the perpetrators manipulated their victims into creating app-specific passwords.

These passwords are unique 16-character codes created by Google which enable secure access to certain applications and devices when two-factor authentication is enabled. As a result of using these app passwords, which bypass conventional two-factor authentication, the attackers were able to gain persistent access to sensitive emails through Gmail accounts undetected.

It is clear from this operation that state-sponsored cyber actors are becoming increasingly inventive, and there is also a persistent risk posed by seemingly secure mechanisms for recovering and accessing accounts. According to Google, this activity was carried out by a threat cluster designated UNC6293, which is closely related to the Russian hacking group known as APT29. It is believed that UNC6293 has been closely linked to APT29, a state-sponsored hacker collective.

APT29 has garnered attention as one of the most sophisticated and sophisticated Advanced Persistent Threat (APT) groups sponsored by the Russian government, and according to intelligence analysts, that group is an extension of the Russian Foreign Intelligence Service (SVR). It is important to note that over the past decade this clandestine collective has orchestrated a number of high-profile cyber-espionage campaigns targeting strategic entities like the U.S. government, NATO member organizations, and prominent research institutes all over the world, including the U.S. government, NATO, and a wide range of academic institutions.

APT29's operators have a reputation for carrying out prolonged infiltration operations that can remain undetected for extended periods of time, characterised by their focus on stealth and persistence. The tradecraft of their hackers is consistently based on refined social engineering techniques that enable them to blend into legitimate communications and exploit the trust of their intended targets through their tradecraft.

By crafting highly convincing narratives and gradually manipulating individuals into compromising security controls in a step-by-step manner, APT29 has demonstrated that it has the ability to bypass even highly sophisticated technical defence systems. This combination of patience, technical expertise, and psychological manipulation has earned the group a reputation as one of the most formidable cyber-espionage threats associated with Russian state interests.

A multitude of names are used by this prolific group in the cybersecurity community, including BlueBravo, Cloaked Ursa, Cosy Bear, CozyLarch, ICECAP, Midnight Blizzard, and The Dukes. In contrast to conventional phishing campaigns, which are based on a sense of urgency or intimidation designed to elicit a quick response, this campaign unfolded in a methodical manner over several weeks.

There was a deliberate approach by the attackers, slowly creating a sense of trust and familiarity with their intended targets. To make their deception more convincing, they distributed phishing emails, which appeared to be official meeting invitations that they crafted. Often, these messages were carefully constructed to appear authentic and often included the “@state.gov” domain as the CC field for at least four fabricated email addresses.

The aim of this tactic was to create a sense of legitimacy around the communication and reduce the likelihood that the recipients would scrutinise it, which in turn increased the chances of the communication being exploited effectively. It has been confirmed that the British writer, Keir Giles, a senior consulting fellow at Chatham House, a renowned global affairs think tank, was a victim of this sophisticated campaign.

A report indicates Giles was involved in a lengthy email correspondence with a person who claimed to be Claudia S Weber, who represented the U.S. Department of State, according to reports. More than ten carefully crafted messages were sent over several weeks, deliberately timed to coincide with Washington's standard business hours. Over time, the attacker gradually gained credibility and trust among the people who sent the messages.

It is worth noting that the emails were sent from legitimate addresses, which were configured so that no delivery errors would occur, which further strengthened the ruse. When this trust was firmly established, the adversary escalated the scheme by sending a six-page PDF document with a cover letter resembling an official State Department letterhead that appeared to be an official State Department document.

As a result of the instructions provided in the document, the target was instructed to access Google's account settings page, to create a 16-character app-specific password labelled "ms.state.gov, and to return the code via email under the guise of completing secure onboarding. As a result of the app password, the threat actors ended up gaining sustained access to the victim's Gmail account, bypassing multi-factor authentication altogether as they were able to access their accounts regularly.

As the Citizen Lab experts were reviewing the emails and PDF at Giles' request, they noted that the emails and PDF were free from subtle language inconsistencies and grammatical errors that are often associated with fraudulent communications. In fact, based on the precision of the language, researchers have suspected that advanced generative AI tools have been deployed to craft polished, credible content for the purpose of evading scrutiny and enhancing the overall effectiveness of the deception as well.

There was a well-planned, incremental strategy behind the attack campaign that was specifically geared towards increasing the likelihood that the targeted targets would cooperate willingly. As one documented instance illustrates, the threat actor tried to entice a leading academic expert to participate in a private online discussion under the pretext of joining a secure State Department forum to obtain his consent.

In order to enable guest access to Google's platform, the victim was instructed to create an app-specific password using Google's account settings. In fact, the attacker used this credential to gain access to the victim's Gmail account with complete control over all multi-factor authentication procedures, enabling them to effectively circumvent all of the measures in place.

According to security researchers, the phishing outreach was carefully crafted to look like a routine, legitimate onboarding process, thus making it more convincing. In addition to the widespread trust that many Americans place in official communications issued by U.S. government institutions, the attackers exploited the general lack of awareness of the dangers of app-specific passwords, as well as their widespread reliance on official communications.

A narrative of official protocol, woven together with professional-sounding language, was a powerful way of making the perpetrators more credible and decreasing the possibility of the target questioning their authenticity in their request. According to cybersecurity experts, several individuals who are at higher risk from this campaign - journalists, policymakers, academics, and researchers - should enrol in Google's Advanced Protection Program (APP).

A major component of this initiative is the restriction of access to only verified applications and devices, which offers enhanced safeguards. The experts also advise organisations that whenever possible, they should disable the use of app-specific passwords and set up robust internal policies that require any unusual or sensitive requests to be verified, especially those originating from reputable institutions or government entities, as well as implement robust internal policies requiring these types of requests.

The intensification of training for personnel most vulnerable to these prolonged social engineering attacks, coupled with the implementation of clear, secure channels for communication between the organisation and its staff, would help prevent the occurrence of similar breaches in the future. As a result of this incident, it serves as an excellent reminder that even mature security ecosystems remain vulnerable to a determined adversary combining psychological manipulation with technical subterfuge when attempting to harm them.

With threat actors continually refining their methods, organisations and individuals must recognise that robust cybersecurity is much more than merely a set of tools or policies. In order to combat cyberattacks as effectively as possible, it is essential to cultivate a culture of vigilance, scepticism, and continuous education. In particular, professionals who routinely take part in sensitive research, diplomatic relations, or public relations should assume they are high-value targets and adopt a proactive defence posture.

Consequently, any unsolicited instructions must be verified by a separate, trusted channel, hardware security keys should be used to supplement authentication, and account settings should be reviewed regularly for unauthorised changes. For their part, institutions should ensure that security protocols are both accessible and clearly communicated as they are technically sound by investing in advanced threat intelligence, simulating sophisticated phishing scenarios, and investing in advanced threat intelligence.

Fundamentally, resilience against state-sponsored cyber-espionage is determined by the ability to plan in advance not only how adversaries are going to deploy their tactics, but also the trust they will exploit in order to reach their goals.

AI Integration Raises Alarms Over Enterprise Data Safety



 

Web Services

AI integration Artificial Intelligence Cyber Crime CyberCrime CyberThreat Cyersecurity AI Data Data Protectoin Technology Google LLM. Okta Web Services

AI Integration Raises Alarms Over Enterprise Data Safety

Today's digital landscape has become increasingly interconnected, and cyber threats have risen in sophistication, which has significantly weakened the effectiveness of traditional security protocols. Cybercriminals have evolved their tactics to exploit emerging vulnerabilities, launch highly targeted attacks, and utilise advanced techniques to breach security perimeters to gain access to and store large amounts of sensitive and mission-critical data, as enterprises continue to generate and store significant volumes of sensitive data.

In light of this rapidly evolving threat environment, organisations are increasingly forced to adopt more adaptive and intelligent security solutions in addition to conventional defences. In the field of cybersecurity, artificial intelligence (AI) has emerged as a significant force, particularly in the area of data protection.

AI-powered data security frameworks are revolutionising the way threats are detected, analysed, and mitigated in real time, making it a transformative force. This solution enhances visibility across complex IT ecosystems, automates threat detection processes, and supports rapid response capabilities by identifying patterns and anomalies that might go unnoticed by human analysts.

Additionally, artificial intelligence-driven systems allow organisations to develop risk mitigation strategies that are scalable as well as aligned with their business objectives while implementing risk-based mitigation strategies. The integration of artificial intelligence plays a crucial role in maintaining regulatory compliance in an era where data protection laws are becoming increasingly stringent, in addition to threat prevention.

By continuously monitoring and assessing cybersecurity postures, artificial intelligence is able to assist businesses in upholding industry standards, minimising operations interruptions, and strengthening stakeholder confidence. Modern enterprises need to recognise that AI-enabled data security is no longer a strategic advantage, but rather a fundamental requirement for safeguarding digital assets in a modern enterprise, as the cyber threat landscape continues to evolve.

Varonis has recently revealed that 99% of organisations have their sensitive data exposed to artificial intelligence systems, a shocking finding that illustrates the importance of data-centric security. There has been a significant increase in the use of artificial intelligence tools in business operations over the past decade. The State of Data Security: Quantifying Artificial Intelligence's Impact on Data Risk presents an in-depth analysis of how misconfigured settings, excessive access rights and neglected security gaps are leaving critical enterprise data vulnerable to AI-driven exploitation.

An important characteristic of this report is that it relies on extensive empirical analysis rather than opinion surveys. In order to evaluate the risk associated with data across 1,000 organisations, Varonis conducted a comprehensive analysis of data across a variety of cloud computing environments, including the use of over 10 billion cloud assets and over 20 petabytes of sensitive data.

Among them were platforms such as Amazon Web Services, Google Cloud Services, Microsoft Azure Services, Microsoft 365 Services, Salesforce, Snowflake, Okta, Databricks, Slack, Zoom, and Box, which provided a broad and realistic picture of enterprise data exposure in the age of Artificial Intelligence. The CEO, President, and Co-Founder of Varonis, Yaaki Faitelson, stressed the importance of balancing innovation with risk, noting that, even though AI is undeniable in increasing productivity, it also poses serious security issues.

Due to the growing pressure on CIOs and CISOs to adopt artificial intelligence technologies at a rapid rate, advanced data security platforms are in increasing demand. It is important to take a proactive, data-oriented approach to cybersecurity to prevent AI from becoming a gateway to large-scale data breaches, says Faitelson. It is important to note that researchers are also exploring two critical dimensions of risk as they relate to large language models (LLMs) as well as AI copilots: human-to-machine interaction and machine-to-machine integrity, which are both critical aspects of risk pertaining to AI-driven data exposure.

A key focus of the study was on how sensitive data, such as employee compensation details, intellectual property rights, proprietary software, and confidential research and development insights able to be unintentionally accessed, leaked, or misused by using just a single prompt into an artificial intelligence interface if it is not protected. As AI assistants are being increasingly used throughout departments, the risk of inadvertently disclosing critical business information has increased considerably.

Additionally, two categories of risk should be addressed: the integrity and trustworthiness of the data used to train or enhance artificial intelligence systems. It is common for machine-to-machine vulnerabilities to arise when flawed, biased, or deliberately manipulated datasets are introduced into the learning cycle of machine learning algorithms.

As a consequence of such corrupted data, it can result in far-reaching and potentially dangerous consequences. For example, inaccurate or falsified clinical information could lead to life-saving medical treatments being developed, while malicious actors may embed harmful code within AI training pipelines, introducing backdoors or vulnerabilities to applications that aren't immediately detected at first.

The dual-risk framework emphasises the importance of tackling artificial intelligence security holistically, one that takes into account the entire lifecycle of data, from acquisition and input to training and deployment, not just the user-level controls. Considering both human-induced and systemic risks associated with generative AI tools, organisations can implement more resilient safeguards to ensure that their most valuable data assets are protected as much as possible.

Organisations should reconsider and go beyond conventional governance models to secure sensitive data in the age of AI. In an environment where AI systems require dynamic, expansive access to vast datasets, traditional approaches to data protection -often rooted in static policies and role-based access -are no longer sufficient.

Towards the future of AI-ready security, a critical balance must be struck between ensuring robust protection against misuse, leakage, and regulatory non-compliance, while simultaneously enabling data access for innovation. Organisations need to adopt a multilayered, forward-thinking security strategy customised for AI ecosystems to meet these challenges.

It is important to note that some key components of a data-tagging and classification strategy are the identification and categorisation of sensitive information to determine how it should be handled depending on the criticality of the information. As a replacement for role-based access control (RBAC), attribute-based access control (ABAC) should allow for more granular access policies based on the identity of the user, context, and the sensitivity of the data.

Aside from that, organisations need to design data pipelines that are AI-aware and incorporate proactive security checkpoints into them so as to monitor how their data is used by artificial intelligence tools. Additionally, output validation becomes crucial—it involves implementing mechanisms that ensure outputs generated by artificial intelligence are compliant, accurate, and potentially risky before they are circulated internally or externally.

The complexity of this landscape has only been compounded by the rise of global regulations and regional regulations that govern data protection and artificial intelligence. In addition to the general data privacy frameworks of GDPR and CCPA, businesses will now need to prepare themselves for emerging AI-specific regulations that will put a stronger emphasis on how AI systems access and process sensitive data. As a result of this regulatory evolution, organisations need to maintain a security posture that is both agile and anticipatable.

Matillion Data Productivity Cloud, for instance, is a solution that embodies this principle of "secure by design". As a hybrid cloud SaaS platform tailored to enterprise environments, Matillion has created a platform that is well-suited to secure enterprise environments.

With its standardised encryption and authentiyoucation protocols, the platform is easily integrated into enterprise networks through the use of a secure cloud infrastructure. This platform is built around a pushdown architecture that prevents customer data from leaving the organisation's own cloud environment while allowing advanced orchestration of complex data workflows in order to minimise the risk of data exposure.

Rather than focusing on data movement, Matillion's focus is on metadata management and workflow automation, providing organisations with a secure, efficient data operation, allowing them to gain insights faster with a higher level of data integrity and compliance. Organisations must move towards a paradigm shift—where security is woven into the fabric of the data lifecycle—as AI poses a dual pressure on organisations.

A shift from traditional governance systems to more adaptive, intelligent frameworks will help secure data in the AI era. Because AI systems require broad access to enterprise data, organisations must strike a balance between openness and security. To achieve this, data can be tagged and classified and attributes can be used to manage access precisely, attribute-based access controls should be implemented for precise control of access, and AI-aware data pipelines must be built with security checks, and output validation must be performed to prevent the distribution of risky or non-compliant AI-generated results.

With the rise of global and AI-specific regulations, companies need to develop compliance strategies that will ensure future success. Matillion Data Productivity Cloud is an example of a platform which offers a secure-by-design solution, as it combines a hybrid SaaS architecture with enterprise-grade security and security controls.

Through its pushdown processing, the customer's data will stay within the organisation's cloud environment while the workflows are orchestrated safely and efficiently. In this way, organisations can make use of AI confidently without sacrificing data security or compliance with the laws and regulations. As artificial intelligence and enterprise data security rapidly evolve, organisations need to adopt a future-oriented mindset that emphasises agility, responsibility, and innovation.

It is no longer possible to rely on reactive cybersecurity; instead, businesses must embrace AI-literate governance models, advance threat intelligence capabilities, and secure infrastructures designed with security in mind. Data security must be embedded into all phases of the data lifecycle, from creation and classification to accessing, analysing, and transforming it with AI. Developing a culture of continuous risk evaluation is a must for leadership teams, and IT and data teams must be empowered to collaborate with compliance, legal, and business units proactively.

In order to maintain trust and accountability, it will be imperative to implement clear policies regarding AI usage, ensure traceability in data workflows, and establish real-time auditability. Further, with the maturation of AI regulations and the increasing demands for compliance across a variety of sectors, forward-looking organisations should begin aligning their operational standards with global best practices rather than waiting for mandatory regulations to be passed.

A key component of artificial intelligence is data, and the protection of that foundation is a strategic imperative as well as a technical obligation. By putting the emphasis on resilient, ethical, and intelligent data security, today's companies will not only mitigate risk but will also be able to reap the full potential of AI tomorrow.

Rust-Developed InfoStealer Extracts Sensitive Data from Chromium-Based Browsers



 

Web browser

Google Internet malware Microsoft RustStealer Web browser

Rust-Developed InfoStealer Extracts Sensitive Data from Chromium-Based Browsers

Browsers at risk

The latest information-stealing malware, made in the Rust programming language, has surfaced as a major danger to users of Chromium-based browsers such as Microsoft Edge, Google Chrome, and others.

Known as “RustStealer” by cybersecurity experts, this advanced malware is made to retrieve sensitive data, including login cookies, browsing history, and credentials, from infected systems.

Evolution of Rust language

The growth in Rust language known for memory safety and performance indicates a transition toward more resilient and hard-to-find problems, as Rust binaries often escape traditional antivirus solutions due to their combined nature and lower order in malware environments.

RustStealers works with high secrecy, using sophisticated obfuscation techniques to escape endpoint security tools. Initial infection vectors hint towards phishing campaigns, where dangerous attachments or links in evidently genuine emails trick users into downloading the payload.

After execution, the malware makes persistence via registry modifications or scheduled tasks, to make sure it remains active even after the system reboots.

Distribution Mechanisms

The main aim is on Chromium-based browsers, abusing the accessibility of unencrypted information stored in browser profiles to harvest session tokens, usernames, and passwords.

Besides this, RustStealer has been found to extract data to remote C2 servers via encrypted communication channels, making detection by network surveillance tools such as Wireshark more challenging.

Experts have also observed its potential to attack cryptocurrency wallet extensions, exposing users to risks in managing digital assets via browser plugins. This multi-faceted approach highlights the malware’s goal to increase data robbery while reducing the chances of early detection, a technique similar to advanced persistent threats (APTs).

About RustStealer malware

What makes RustStealer different is its modular build, letting hackers rework its strengths remotely. This flexibility reveals that future ve

This adaptability suggests that future replications could integrate functionalities such as ransomware components or keylogging, intensifying threats in the longer run.

The deployment of Rust also makes reverse-engineering efforts difficult, as the language’s output is less direct to decompile in comparison to scripts like Python or other languages deployed in outdated malware strains.

Businesses are advised to remain cautious, using strong phishing securities, frequently updating browser software, and using endpoint detection and response (EDR) solutions to detect suspicious behavior.

APT41 Exploits Google Calendar in Stealthy Cyberattack; Google Shuts It Down



 

threat

APT41 Attack Calendar Cybersecurity Detection Google Hacking malware phishing threat

APT41 Exploits Google Calendar in Stealthy Cyberattack; Google Shuts It Down

Chinese state-backed threat actor APT41 has been discovered leveraging Google Calendar as a command-and-control (C2) channel in a sophisticated cyber campaign, according to Google’s Threat Intelligence Group (TIG). The team has since dismantled the infrastructure and implemented defenses to block similar future exploits.

The campaign began with a previously breached government website — though TIG didn’t disclose how it was compromised — which hosted a ZIP archive. This file was distributed to targets via phishing emails.

Once downloaded, the archive revealed three components: an executable file and a dynamic-link library (DLL) disguised as image files, and a Windows shortcut (LNK) masquerading as a PDF. When users attempted to open the phony PDF, the shortcut activated the DLL, which then decrypted and launched a third file containing the actual malware, dubbed ToughProgress.

Upon execution, ToughProgress connected to Google Calendar to retrieve its instructions, embedded within event descriptions or hidden calendar events. The malware then exfiltrated stolen data by creating a zero-minute calendar event on May 30, embedding the encrypted information within the event's description field.

Google noted that the malware’s stealth — avoiding traditional file installation and using a legitimate Google service for communication — made it difficult for many security tools to detect.

To mitigate the threat, TIG crafted specific detection signatures, disabled the threat actor’s associated Workspace accounts and calendar entries, updated file recognition tools, and expanded its Safe Browsing blocklist to include malicious domains and URLs linked to the attack.

Several organizations were reportedly targeted. “In partnership with Mandiant Consulting, GTIG notified the compromised organizations,” Google stated. “We provided the notified organizations with a sample of TOUGHPROGRESS network traffic logs, and information about the threat actor, to aid with detection and incident response.”

Google did not disclose the exact number of impacted entities.

Google’s AI Virtual Try-On Tool Redefines Online Shopping Experience



 

Privacy

AI Systems AI technology AI tools customer data privacy Data Privacy Data Safety User Data data security Data Usage Google Google AI model Privacy

Google’s AI Virtual Try-On Tool Redefines Online Shopping Experience

At the latest Google I/O developers conference, the tech giant introduced an unexpected innovation in online shopping: an AI-powered virtual try-on tool. This new feature lets users upload a photo of themselves and see how clothing items would appear on their body. By merging the image of the user with that of the garment, Google’s custom-built image generation model creates a realistic simulation of the outfit on the individual.

While the concept seems simple, the underlying AI technology is advanced. In a live demonstration, the tool appeared to function seamlessly. The feature is now available in the United States and is part of Google’s broader efforts to enhance the online shopping experience through AI integration. It’s particularly useful for people who often struggle to visualize how clothing will look on their body compared to how it appears on models.

However, the rollout of this tool raised valid questions about user privacy. AI systems that involve personal images often come with concerns over data usage. Addressing these worries, a Google representative clarified that uploaded photos are used exclusively for the try-on experience. The images are not stored for AI training, are not shared with other services or third parties, and users can delete or update their photos at any time. This level of privacy protection is notable in an industry where user data is typically leveraged to improve algorithms.

Given Google’s ongoing development of AI-driven tools, some expected the company to utilize this photo data for model training. Instead, the commitment to user privacy in this case suggests a more responsible approach. Virtual fitting technology isn’t entirely new. Retail and tech companies have been exploring similar ideas for years. Amazon, for instance, has experimented with AI tools in its fashion division. Google, however, claims its new tool offers a more in-depth understanding of diverse body types.

During the presentation, Vidhya Srinivasan, Google’s VP of ads and commerce, emphasized the system’s goal of accommodating different shapes and sizes more effectively. Past AI image tools have faced criticism for lacking diversity and realism. It’s unclear whether Google’s new tool will be more reliable across the board. Nevertheless, their assurance that user images won’t be used to train models helps build trust.

Although the virtual preview may not always perfectly reflect real-life appearances, this development points to a promising direction for AI in retail. If successful, it could improve customer satisfaction, reduce returns, and make online shopping a more personalized experience.

Google Unveils AI With Deep Reasoning and Creative Video Capabilities



 

Software

Advanced Technology Artificial Intelligence Cyber Security cybercriminals CyberThreat Google Google Ads Google Marketing Software

Google Unveils AI With Deep Reasoning and Creative Video Capabilities

This week, Google, as part of its annual Google Marketing Live 2025 event, unveiled a comprehensive suite of artificial intelligence-powered tools to help the company cement its position at the forefront of digital commerce and advertising on Wednesday, May 21, at a press conference.

Google's new tools are intended to revolutionise the way brands engage with consumers and drive measurable growth through artificial intelligence, and they are part of a strategic push that Google is making to redefine the future of advertising and online shopping. In her presentation, Vidhya Srinivasan, Vice President and General Manager of Google Ads and Commerce, stressed the importance of this change, saying, “The future of advertising is already here, fueled by artificial intelligence.”

This declaration was followed by Google's announcement of advanced solutions that will enable businesses to use smarter bidding, dynamic creative creation, and intelligent, agent-based assistants in real-time, which can adjust to user behaviour and market conditions, as well as adapt to changing market conditions. Google has launched this major product at a critical time in its history, as generative AI platforms and conversational search tools are putting unprecedented pressure on traditional search and shopping channels, diverting users away from these methods.

By leveraging technological disruptions as an opportunity for brands and marketers around the world, Google underscores its commitment to staying ahead of the curve by creating innovation-driven opportunities for brands and marketers. A long time ago, Google began to explore artificial intelligence, and since its inception in 1998, it has evolved steadily. Google’s journey into artificial intelligence dates back much earlier than many people think.

While Google has always been known for its groundbreaking PageRank algorithm, its formal commitment to artificial intelligence accelerated throughout the mid-2000s when key milestones like the acquisition of Pyra Labs in 2003 and the launch of Google Translate in 2006 were key milestones. It is these early efforts that laid the foundation for analysing content and translating it using AI. It was not long before Google Instant was introduced in 2010 as an example of how predictive algorithms were enhancing user experience by providing real-time search query suggestions.

In the years that followed, artificial intelligence research and innovation became increasingly important, as evidenced by Google X's establishment in 2011 and DeepMind's strategic acquisition in 2014, pioneers in reinforcement learning that created the historic algorithm AlphaGo. A new wave of artificial intelligence has been sweeping across the globe since 2016 with Google Assistant and advanced tools like TensorFlow, which have democratized machine learning development.

Breakthroughs such as Duplex have highlighted AI's increasing conversational sophistication, but most recently, Google's AI has embraced multimodal capabilities, which is why models like BERT, LaMDA, and PaLM are revolutionising language understanding and dialogue in a way previously unknown to the world. AI has a rich legacy that underscores its crucial role in driving Google’s transformation across search, creativity, and business solutions, underpinned by this legacy.

As part of its annual developer conference in 2025, Google I/O reaffirmed its leadership in the rapidly developing field of artificial intelligence by unveiling an impressive lineup of innovations that promise to revolutionize the way people interact with technology, reaffirming its leadership in this field. In addition to putting a heavy emphasis on artificial intelligence-driven transformation, this year's event showcased next-generation models and tools that are far superior to the ones displayed in previous years.

Among the announcements made by AI are the addition of AI assistants with deeper contextual intelligence, to the creation of entire videos with dialogue, which highlights a monumental leap forward in both the creative and cognitive capabilities of AI in general. It was this technological display that was most highlighted by the unveiling of Gemini 2.5, Google's most advanced artificial intelligence model. This model is positioned as the flagship model of the Gemini series, setting new industry standards for outstanding performance across key dimensions, such as reasoning, speed, and contextual awareness, which is among the most important elements of the model.

The Gemini 2.5 model has outperformed its predecessors and rivals, including Google's own Gemini Flash, which has redefined expectations for what artificial intelligence can do. Among the model's most significant advantages is its enhanced problem-solving ability, which makes it far more than just a tool for retrieving information; it is also a true cognitive assistant because it provides precise, contextually-aware responses to complex and layered queries.

It has significantly enhanced capabilities, but it operates at a faster pace and with better efficiency, which makes it easier to integrate into real-time applications, from customer support to high-level planning tools, seamlessly. Additionally, the model's advanced understanding of contextual cues allows it to conduct intelligent, more coherent conversations, allowing it to feel more like a human being collaborating rather than interacting with a machine. This development marks a paradigm shift in artificial intelligence in addition to incremental improvements.

It is a sign that artificial intelligence is moving toward a point where systems are capable of reasoning, adapting, and contributing in meaningful ways across the creative, technical, and commercial spheres. Google I/O 2025 serves as a preview of a future where AI will become an integral part of productivity, innovation, and experience design for digital creators, businesses, and developers alike.

Google has announced that it is adding major improvements to its Gemini large language model lineup, which marks another major step forward in Google's quest to develop more powerful, adaptive artificial intelligence systems, building on the momentum of its breakthroughs in artificial intelligence. The new iterations, Gemini 2.5 Flash and Gemini 2.5 Pro, feature significant architectural improvements that aim to optimise performance across a wide range of uses.

It will be available in early June 2025 in general availability as Gemini 2.5 Flash, a fast and lightweight processor designed for high-speed and lightweight use, and the more advanced Pro version will appear shortly afterwards as well. Among the most notable features of the Pro model is the introduction of “Deep Think” which provides advanced reasoning techniques to handle complex tasks using parallel processing techniques to handle complex issues.

As a result of its inspiration from AlphaGo's strategic modelling, Deep Think gives AI the ability to simultaneously explore various solution paths, producing faster and more accurate results. With this capability, the model is well-positioned to offer a cutting-edge solution for reasoning at the highest level, mathematical analysis, and programming that meets the demands of competition. When Demiss Hassabis, CEO of Google DeepMind, held a press briefing to highlight the model's breakthrough performance, he highlighted its impressive performance on the USAMO 2025, a challenging math challenge that is a challenging one in the world, and LiveCodeBench, another benchmark that is a popular one in advanced coding.

A statement by Hassabis said, “Deep Think pushed the performance of models to the limit, resulting in groundbreaking results.” Google is adopting a cautious release strategy to comply with its commitment to ethical AI deployment. In order to ensure safety, reliability, and transparency, Deep Think will initially be accessible only to a limited number of trusted testers who will be able to provide feedback.

In addition to demonstrating Google's intent to responsibly scale frontier AI capabilities, this deliberate rollout emphasises the importance of maintaining trust and control while showcasing the company's commitment to it. In addition to its creative AI capabilities, Google announced two powerful models for generative media during its latest announcements: Veo 3 for video generation and Imagen 4. These models represent significant breakthroughs in generative media technology.

There has been a shift in artificial intelligence-assisted content creation in recent years, and these innovations provide creators with a much deeper, more immersive toolkit that allows them to tell visual and audio stories in a way that is truly remarkable in terms of realism and precision. Veo 3 represents a transformative leap in video generation technology, and for the first time, artificial intelligence-generated videos do not only comprise silent, motion-only clips anymore, but also provide a wide range of visual effects and effects.

As a result of the integration of fully synchronised audio with Veo 3, the experience felt more like a real cinematic production than a simple algorithmic output, with ambient sounds, sound effects, and even real-time dialogue between characters, as it was in the original film. "For the first time in history, we are entering into a new era of video creation," said Demis Hassabis, CEO of Google DeepMind, highlighting how both the visual fidelity and the auditory depth of the new model were highlighted. As a result of these breakthroughs, Google has developed Flow, a new AI-powered filmmaking platform exclusively for creative professionals, which integrates these breakthroughs into Flow.

Flow is Google's latest generative modelling tool that combines the most advanced models into an intuitive interface, so storytellers can design cinematic sequences with greater ease and fluidity than ever before. In Flow, the company claims it will recreate the intuitive, inspired creative process, where iteration feels effortless and ideas evolve in a way that is effortless and effortless. Flow has already been used by several filmmakers to create short films that illustrate the creative potential of the technology, combining Flow's capabilities with traditional methods to create the films.

Additionally, Imagen 4 is the latest update to Google's image generation model, offering extraordinary improvements in visual clarity, fine detail, and especially in typography and text rendering, as well as providing unparalleled advancements in visual clarity and fine detail. With these improvements, it has become a powerful tool for marketers, designers, and content creators who need to create beautiful visuals combining high-quality imagery with precise, readable text.

The Imagen 4 platform is a significant step forward in advancing the quality of visual storytelling based on artificial intelligence, whether for branding, digital campaigns, or presentations. Despite fierce competition from leading technology companies, Google has made significant advancements in autonomous artificial intelligence agents at a time when the landscape of intelligent automation is rapidly evolving.

It is no secret that Microsoft's GitHub Copilot has already demonstrated how powerful AI-driven development assistants can be, but OpenAI's CodeX platform continues to push the boundaries of what AI has to offer. It is in this context that Google introduced innovative tools like Stitch and Jules that could generate a complete website, a codebase, and a user interface automatically without any human input. These tools signal a revolution in how software developers develop and create digital content. A convergence of autonomous artificial intelligence technologies from a variety of industry giants underscores a trend towards automating increasingly complex knowledge tasks.

Through the use of these AI systemorganisationsons can respond quickly to changing market demands and evolving consumer preferences by providing real-time recommendations and dynamic adjustments. Through such responsiveness, an organisation is able to optimise operational efficiency, maximise resource utilisation, and create sustainable growth by ensuring that the company remains tightly aligned with its strategic goals. AI provides businesses with actionable insights that enable them to compete more effectively in an increasingly complex and fast-paced market place by providing actionable insights.

Aside from software and business applications, Google's AI innovations also have great potential to have a dramatic impact on the healthcare sector, where advancements in diagnostic accuracy and personalised treatment planning have the potential to greatly improve the outcomes for patients. Furthermore, improvements in the field of natural language processing and multimodal interaction models will help provide more intuitive, accessible and useful user interfaces for users from diverse backgrounds, thus reducing barriers to adoption and enabling them to make the most of technology.

In the future, when artificial intelligence becomes an integral part of today's everyday lives, its influence will be transformative, affecting industries, redefining workflows, and generating profound social effects. The fact that Google leads the way in this space not only implies a future where artificial intelligence will augment human capabilities, but it also signals the arrival of a new era of progress in science, economics, and culture as a whole.

Search This Blog

Sections

Popular Posts

Blog Archive

Labels

Report Abuse

About Me

Showing result(s) for

Popular Posts

Pages

Google Gemini Exploit Enables Covert Delivery of Phishing Content

Google Gemini Bug Exploits Summaries for Phishing Scams

False AI summaries leading to phishing attacks

Gemini for attack

Impact

Security Breach Reveals "Catwatchful" Spyware is Snooping on Users

About Catwatchful

Rise of spyware apps

How was the spyware found?

Ditch Passwords, Use Passkeys to Secure Your Account

Ditch passwords, use passkeys

What are passkeys?

Okta warns users to not use passwords

Why are passwords not safe?

New Android Feature Detects Fake Mobile Networks

WhatsApp Under Fire for AI Update Disrupting Group Communication

OpenAI Rolls Out Premium Data Connections for ChatGPT Users

Russian Threat Actors Circumvent Gmail Security with App Password Theft

AI Integration Raises Alarms Over Enterprise Data Safety

Rust-Developed InfoStealer Extracts Sensitive Data from Chromium-Based Browsers

Browsers at risk

Evolution of Rust language

Distribution Mechanisms

About RustStealer malware

APT41 Exploits Google Calendar in Stealthy Cyberattack; Google Shuts It Down

Google’s AI Virtual Try-On Tool Redefines Online Shopping Experience

Google Unveils AI With Deep Reasoning and Creative Video Capabilities

Footer About

Search This Blog

Sections

Popular Posts

Blog Archive

Labels

About Me

Showing result(s) for

Popular Posts

Pages

Menu Item

False AI summaries leading to phishing attacks

Gemini for attack

Impact

About Catwatchful

Rise of spyware apps

How was the spyware found?

Ditch passwords, use passkeys

What are passkeys?

Okta warns users to not use passwords

Why are passwords not safe?

Browsers at risk

Evolution of Rust language

Distribution Mechanisms

About RustStealer malware