Search This Blog
Powered by Blogger.

Blog Archive

Labels

Footer About

Footer About

Labels

Load More
Trendy Right Now

Popular Posts

Cyber Security
Ad Data Privacy AI Chatbot AI Privacy AI Security chatbot security Cyber Privacy Cyber Security

WhatsApp Incognito AI Chats Raise Privacy and Accountability Concerns

 

Private AI chats are now arriving on WhatsApp through a new incognito mode where conversations disappear once they end. Neither users nor Meta will retain copies of these exchanges, according to the company. Executives say the feature was designed for sensitive discussions involving health, finances, relationships, or personal struggles, where users may not want permanent records stored online. 

Unlike most AI systems that retain chat history for moderation, improvements, or future model training, Meta claims these AI conversations will not be saved on company servers at all. CEO Mark Zuckerberg described it as one of the first major AI systems built without maintaining conversation logs. According to Will Cathcart, many users feel uncomfortable sharing private information when companies can later review chat histories. 

To address this, the new setting automatically erases AI discussions after completion, leaving no retrievable record behind. Although WhatsApp says the feature provides protections similar to end-to-end encryption, the company acknowledged the underlying technology differs from the encryption used for regular WhatsApp messages. Meta nevertheless maintains that users should expect comparable privacy safeguards while interacting with AI tools. 

Despite the stronger privacy focus, cybersecurity experts warn the system could create accountability challenges. Alan Woodward from the University of Surrey noted that while the feature is unlikely to weaken WhatsApp’s broader security infrastructure, disappearing AI chats could make it difficult to investigate harmful responses or dangerous recommendations generated by the chatbot. Companies including OpenAI and Google have already faced legal scrutiny tied to allegations that AI conversations contributed to emotional harm, unsafe behavior, or psychological distress. 

If AI chats vanish permanently, neither users nor Meta may be able to review what was said during critical interactions. Experts also warn that disappearing chat histories may reduce transparency around misinformation, moderation failures, or unsafe advice shared privately by AI systems. Without stored records, proving what responses were generated during sensitive moments becomes far more difficult. Meta says additional safety protections are still being developed. 

Initially, the incognito mode will support only text conversations rather than image processing, while stricter moderation guardrails are expected to block prompts considered harmful, illegal, or dangerous. The feature also reflects Meta’s broader push to integrate AI across Instagram, Facebook, and Messenger. Despite criticism from some users after Meta AI was added to WhatsApp without a full removal option, the company continues aggressively expanding its AI ecosystem. 

Industry analysts say Meta’s growing investment in AI infrastructure is tied to intense competition across the technology sector. The company is expected to spend heavily on artificial intelligence throughout 2026 to improve advertising systems, shopping features, and user engagement tools. Investors, however, remain cautious about whether those enormous investments will ultimately generate long-term returns. 

WhatsApp’s disappearing AI conversations highlight an increasingly important debate surrounding privacy and accountability. While users may value confidential AI interactions, experts warn that removing all conversation records could also make it harder to investigate misuse, harmful outcomes, or dangerous AI behavior later on.
Read more »
User Privacy
Cyber Security Meta Meta Glass Smart Glasses User Privacy

Meta Smart Glasses Secretly Film Women: Privacy Invasion Crisis Explained

 

Smart glasses are moving from novelty to mainstream, and Meta’s Ray-Ban model is leading the market. The BBC says Meta accounts for about 80% of sales in the smart-glasses category, helped by the familiar Ray-Ban design and the addition of a built-in camera, speakers, and AI features. That combination has made the product appealing to early adopters who want hands-free music, calls, photos, and information on the go. 

But the same features that make smart glasses attractive also make them controversial. The report describes women being filmed without their knowledge by men wearing the glasses, often in everyday settings such as beaches, shops, and sidewalks. Those videos can later appear online and attract harassment, while the people recorded may not even realize it happened until much later. 

Privacy concerns are not limited to casual misuse. The report says some wearers have been surprised to discover what their glasses were recording, while lawsuits have also been filed over videos captured through the devices and used for AI training. In addition, experts quoted in the report warn that if smart glasses become common, it may become much harder to enforce norms around sensitive places like courthouses, hospitals, museums, and bathrooms. 

Meta says the glasses are designed with privacy in mind and that users should behave responsibly. The company’s spokesperson told the BBC that it has teams focused on limiting misuse, but also argued that the ultimate responsibility lies with individual users. Even so, the report notes that visible indicators like the recording light may be too subtle to reliably alert bystanders, especially in bright outdoor conditions.

Despite the backlash, the commercial momentum is strong, and other major tech firms are preparing their own versions. Apple, Snap, and Google are all reportedly working on smart-glasses products, suggesting this could become a major new consumer category rather than a passing trend. The BBC’s reporting points to a familiar tech dilemma: a device can be genuinely useful while still raising difficult questions about consent, surveillance, and the limits of public privacy.
Read more »
WhatsApp Scam
Android Malware APK Malware Cyber Attacks Cyber Fraud Digital Invitation Scam Mobile Banking Threats OTP Theft UPI Fraud WhatsApp Scam

Rising Digital Invitation Scams Highlight Need for Strong Cyber Awareness


 

What was once used for birthdays, weddings, corporate events, and social gatherings has increasingly been weaponized by cybercriminals as a sophisticated phishing technique. 

The security research community has observed that threat actors are increasingly using commonly used invitation platforms and compromised email accounts to distribute fraudulent event links designed to harvest credential information, financial data, and sensitive personal information by leveraging their credibility.

It is evident how even routine online interactions are becoming part of the modern cyber threat landscape when malicious emails mimic legitimate invitation services and utilize the psychological urgency of social engagement. This highlights how even routine online interactions are now a source of cyber threats. 

A cybersecurity investigator has noted that the threat is now extending far beyond deceptive email invitations, as hackers are actively distributing malware-laced Android Package Kit (APK) files disguised as digital event invitations via messaging platforms such as WhatsApp and Telegram. 

A malicious file is often accompanied by socially engineered labels, such as wedding invitations, housewarming ceremonies, or private party invitations, which are designed to reduce suspicion and stimulate immediate downloads. It often mimics utility tools, but remains operationally dormant to avoid detection once installed on an Android device. 

Once embedded, the rogue application quietly embeds itself among legitimate applications, frequently imitating utility tools. It has been reported that victims unknowingly grant extensive permissions to threat actors, including access to call logs, SMS services, notifications, contacts, and screen recording capabilities, effectively giving them deep surveillance access to their devices.

Several observed cases have demonstrated that the malware can intercept one-time passwords, monitor banking and UPI sessions in real-time, and harvest financial credentials directly from user screen activity. Recently, a Bengaluru-based business owner has experienced the severity of the attack chain after receiving a fraudulent wedding invitation APK through WhatsApp, causing unauthorized access to financial information and a financial loss of approximately 5 lakh before detection of the compromise. 

A number of researchers investigating these campaigns have concluded that the attack infrastructure is typically conducted using two highly effective compromise methods that bypass user suspicion and device-level trust mechanisms. As a result of interaction with the malicious invitation link, the link appears broken or inactive. However, behind-the-scenes processes silently deploy credential-stealing malware that harvests passwords, device information, and sensitive personal information. 

Secondly, victims are directed to convincingly spoofed login portals in which their account credentials are captured in real time, allowing threat actors access to banking, email, and payment services without their consent. 

A number of fraudulent invitations deliberately avoid detailed event information in order to induce impulsive clicks, depending instead on urgency and familiarity. In addition to users being advised to treat unsolicited invitations with caution, particularly those received through messaging applications or from unknown senders, IT security experts also recommend reporting and deleting suspicious e-mails as soon as they become aware of them. 

According to threat intelligence firm CloudSEK, these campaigns have resulted in large-scale financial fraud operations. Within 48 hours, one threat group processed transactions worth nearly 25-30,000 crores, emphasizing the rapid scalability of the ecosystem and the high number of victims involved. Specifically, the firm found that the attacks exploit the trust architecture behind SIM-based verification systems commonly used by UPI platforms. 

In such systems, device-linked mobile numbers are considered proof of legitimate account ownership. A malicious APK disguised as a traffic violation notice or a digital invitation is often the first step in establishing covert access to a smartphone's messaging features after securing SMS permissions. 

After deploying the so-called “Digital Lutera” toolkit, CloudSEK indicated that attackers manipulate identity validations and SMS workflows through a specialized Android framework on separate devices. 

With this feature, bank registration messages may be intercepted and OTPs are silently forwarded to attacker-controlled Telegram channels without the victim's knowledge. Additionally, the report revealed that fabricated "sent" SMS records are inserted into message histories in order to maintain an illusion of legitimate activity, such that UPI applications are misled into believing that authentication requests originate from the victim's own smartphone.

Thus, cybercriminals have the opportunity to remotely register and manage the UPI account of a victim even when the original SIM card remains physically in the user's possession. Previously, CloudSEK notified regulators and financial institutions in order to strengthen mitigation frameworks before the threat expands. As part of its responsible disclosure process, it said that it has already notified regulators and financial institutions. 

The convergence of digital payment ecosystems and mobile-first communication platforms represents a shift toward socially engineered, device-centric financial attacks, warn cybersecurity experts. Threat actors are increasingly exploiting human behavior and weaknesses in authentication workflows to exploit APK sideloading, SMS intercept frameworks, and compromised messaging channels as a means of exploiting trust-driven human behaviour.

A stronger understanding of user awareness, stricter application permission controls, and enhanced anomaly detection across UPI and telecommunication infrastructure will assist in limiting the operational scale of these fraud networks before they become a more persistent threat to India's rapidly expanding digital sector.
Read more »
technology
Artificial Intelligence Claude GitHub Opencode technology

OpenCode’s Rapid Growth Reflects Rising Developer Concerns Over AI Vendor Dependence

 





A glaring divide is emerging in the AI coding industry as developers increasingly weigh the convenience of fully managed coding platforms against the flexibility of open-source alternatives designed to avoid dependence on a single provider.

The debate intensified this week after Anthropic used its first “Code with Claude” developer conference to showcase major upgrades across its Claude Code ecosystem. The company announced that rate limits for Claude Code users on Pro, Max, Team, and Enterprise plans would be significantly expanded, while peak-hour usage restrictions were removed entirely. Anthropic also raised usage limits for its Opus API and disclosed a major infrastructure agreement with SpaceX involving the Colossus 1 data center.

According to the company, the agreement will provide access to more than 300 megawatts of computing power and approximately 220,000 Nvidia GPUs expected to come online within weeks. The move reflects the broader AI industry race to secure high-performance computing infrastructure as demand for generative AI services continues to increase.

Anthropic also introduced several updates aimed at turning Claude Code into a more advanced managed development environment. These included expanded Managed Agents capabilities, support for coordinating multiple AI agents simultaneously, a public beta feature called Outcomes, and an experimental memory system internally referred to as “dreaming,” which is intended to help AI systems retain and improve contextual understanding over time.

During the event, Anthropic executive Boris Cherny demonstrated remote agents and automated routines capable of running coding tasks asynchronously, effectively allowing Claude Code to function more like a workflow orchestration platform rather than a traditional coding assistant.

At the same time, a separate trend has been accelerating across the open-source community. OpenCode, an independent coding harness project associated with SST, has experienced a dramatic rise in popularity after positioning itself as an alternative to vendor-controlled AI development environments.

The project’s GitHub repository has now surpassed 157,000 stars, overtaking the roughly 122,000 stars associated with Anthropic’s own Claude Code repository at the time of reporting. While GitHub stars do not necessarily represent active users or production deployments, they are often viewed as indicators of developer awareness, interest, and community support.

The roots of OpenCode’s instant growth trace back to January 2026, when Anthropic introduced server-side authentication checks that prevented third-party tools from accessing Claude Pro and Max subscriptions through OAuth-based authentication methods.

Several projects, including OpenCode, Cline, and RooCode, were affected by the policy change. Prior to the restrictions, these tools allowed developers to run autonomous coding workflows through fixed-price Claude subscriptions rather than paying significantly higher API-based usage fees tied to token consumption.

From Anthropic’s perspective, the restriction addressed a business and infrastructure problem. Subscription plans were designed to support usage within the company’s own ecosystem, while third-party tools were effectively redirecting high-volume workloads through pricing structures never intended for external automation platforms.

Discussions across developer forums, including lengthy conversations on Hacker News, showed that many users understood Anthropic’s reasoning. However, criticism quickly emerged over the manner in which the restrictions were enforced. Developers reported that the changes were introduced without advance notice, disrupting workflows in active sessions. Some users also claimed that automated abuse-detection systems temporarily restricted accounts during the transition period.

OpenCode responded rapidly after the restrictions took effect. The project added support for ChatGPT Plus integrations within hours and began expanding compatibility across multiple AI providers. Anthropic later formalized its position in updated Terms of Service published in February, clarifying that subscription OAuth tokens were not intended for third-party routing or automation tools.

The dispute escalated further in March after OpenCode reportedly received legal requests related to Claude subscription authentication. Shortly afterward, the project merged an update removing references to Claude Pro and Max authentication from its codebase. By April 4, Anthropic’s enforcement measures had expanded to additional third-party harnesses, including OpenClaw and NanoClaw, pushing developers toward pay-as-you-go API billing structures.

Interest in OpenCode accelerated during this period. On March 21, a Hacker News discussion surrounding the project gained more than 1,200 points and hundreds of comments, driving additional visibility across the developer community. By early April, the repository had already crossed 120,000 GitHub stars.

As of May 8, project activity data showed approximately 156,904 stars, 18,259 forks, 4,788 issues, and more than 1,600 open pull requests. OpenCode’s website also claimed participation from over 850 contributors and estimated usage among roughly 6.5 million monthly developers.

Industry observers note that the OAuth dispute alone likely does not explain OpenCode’s growth. Instead, the incident appears to have accelerated an existing movement toward model-agnostic development tools. OpenCode gradually shifted its messaging away from low-cost Claude access and toward provider neutrality, emphasizing that developers should be able to switch between AI models as pricing, performance, and capabilities evolve.

That distinction is increasingly important as competition intensifies between major AI providers. A developer using a model-agnostic harness can move between Anthropic, OpenAI, or other models with relatively minor configuration changes. In contrast, developers operating entirely within a vertically integrated ecosystem may face higher switching costs if pricing structures, usage limits, or platform policies change unexpectedly.

The debate mirrors earlier divisions within the software infrastructure industry. Some analysts have compared the current situation to Docker and Podman, where one platform focused heavily on integrated services and managed workflows while the other prioritized portability, operational control, and independence from platform lock-in.

OpenCode’s rise has also drawn criticism from parts of the developer community. Users in public discussions have raised concerns about high memory usage, the growing complexity of the project’s TypeScript codebase, inconsistent release stability, and the broader security implications of integrating multiple AI providers into a single framework.

Security considerations remain particularly relevant because every additional provider connection potentially expands the software’s attack surface. OpenCode also faced backlash after removing Claude subscription authentication support following reported legal pressure, with some developers expressing frustration over how the project handled the situation.

Still, the overall ndustry direction appears increasingly clear. Anthropic is investing heavily in a future built around tightly managed AI coding ecosystems that combine infrastructure, orchestration, memory systems, and coding assistance within a single platform.

At the same time, open-source projects such as OpenCode, Cline, Aider, and OpenClaw continue to attract developers seeking portability and reduced dependency on individual AI vendors.

For many software teams, the central issue is no longer choosing between Claude Code and OpenCode alone. Instead, developers are beginning to decide whether critical AI-assisted workflows should remain under the control of a single provider or operate through more flexible systems capable of adapting as the AI landscape continues to shift.

Read more »
Data Theft
Cyber Attacks Cyber Data Cyber Security Ransomware Attacks CyberCriminal Data protection data security Data Theft

Foxconn Cyberattack Exposes Alleged Intel, Apple, Nvidia and Google Project Data

 

A wave of digital intrusion lately hit Foxconn, causing interruptions across certain segments of its North American facilities when the Nitrogen ransomware collective admitted involvement - disclosing they had infiltrated systems and extracted vast troves of confidential information. This incident underscores, yet again, how intensifying demands from cybercriminal networks now challenge critical links within international tech logistics, particularly those manufacturers embedded deep inside the production ecosystems serving top-tier technology brands. 

Later on, after initial reports emerged, Foxconn confirmed disruptions across multiple sites in North America. Right away, its cyber defense units began executing crisis protocols instead of waiting for further escalation. Because systems required immediate protection, temporary measures went into place to shield manufacturing flow. Even so, certain plants experienced brief halts in daily activity due to digital interference. Gradually now, output levels are stabilizing following those earlier setbacks. 

Later, the ransomware operators listed Foxconn on their public leak page, stating they had taken close to 8 terabytes of data - over 11 million individual files. Their claim centers on possession of private technical records: blueprints, project directives meant for internal use, engineering schematics. Information tied to big tech names like Apple, Nvidia, Intel, Google, and Dell reportedly appears within what was pulled. Though unverified, the alleged haul suggests access to development assets considered highly sensitive. 

Even though hackers say they took customer data, Foxconn hasn’t said if any was truly exposed. Without a clear statement, it remains unclear how much information may have been reached - or if partner details were touched at all. Ever since 2023, the Nitrogen ransomware crew has operated under suspicion of ties to variants spawned from exposed Conti 2 code. Researchers point out weaknesses in their tools - especially when striking VMware ESXi systems. 

Despite handing over payments, certain targets still could not retrieve locked data. This failure stems from defective decryption mechanisms built directly into the malicious software. Recovery gaps appear baked into its flawed design. Should that glitch persist, affected groups might face deeper troubles - offering money to hackers does not always bring back locked data or recover what was taken. Back in 2024, the LockBit group took credit for breaching Foxsemicon Integrated Technology - a firm within the larger Foxconn Technology Group. 

It wasn’t an isolated case; a similar unit of Foxconn in Mexico had drawn their attention two years prior. Ransomware attacks on this network are nothing new. The pattern stretches further back than it might first appear. Now worries spread through the hardware world after the recent security incident, given how central Foxconn is to building devices and moving parts for big tech firms worldwide. 

When something interferes with its work, delays may ripple into assembly timelines, logistics systems, operational frameworks, even sensitive processes behind upcoming gadgets and corporate tools. Because they rely on many partners, handle valuable technical details, and face tight deadlines when operations fail, factories and logistics companies often attract ransomware groups. 

With more strikes hitting essential vendors lately, better separation between internal systems is becoming a priority - alongside stronger crisis plans and tighter protection for confidential design files that could be stolen or leaked.
Read more »
zero-day vulnerability
AI-generated exploit Artificial Intelligence Cybersecurity Google Threat Intelligence Group Technology zero-day vulnerability

Google Detects AI-Generated Zero-Day Exploit Targeting Web Admin Tool

 

Researchers from Google Threat Intelligence Group (GTIG) have revealed that a recently identified zero-day exploit aimed at a widely used open-source web administration platform was likely created with the help of artificial intelligence.

The vulnerability, which targeted the platform’s two-factor authentication (2FA) mechanism, could have allowed attackers to bypass critical security protections. While the software involved has not been publicly identified, researchers confirmed that the attack was stopped before it reached large-scale exploitation.

According to GTIG, analysis of the Python-based exploit strongly indicates the involvement of AI tools during the vulnerability discovery and weaponization process. The team noted that the coding style, educational explanations within the script, and even fabricated technical details closely resembled outputs commonly produced by large language models (LLMs).

“For example, the script contains an abundance of educational docstrings, including a hallucinated CVSS score, and uses a structured, textbook Pythonic format highly characteristic of LLMs training data,” GTIG says in a report today.

Researchers also stated that the flaw itself appeared to be a semantic logic issue — an area where AI systems tend to perform effectively — rather than traditional vulnerabilities like memory corruption or poor input sanitization that are usually identified through fuzzing or static analysis techniques.

Google informed the affected software developer about the issue, allowing security measures to be implemented quickly and the attack to be disrupted before wider abuse occurred.

“For the first time, GTIG has identified a threat actor using a zero-day exploit that we believe was developed with AI,” GTIG researchers say.

The report additionally highlights the increasing role of AI in cybercrime operations. Google observed threat groups linked to China and North Korea — including APT27, APT45, UNC2814, UNC5673, and UNC6201 — using AI systems for exploit development and vulnerability research.

Meanwhile, Russia-associated threat actors were reportedly using AI-generated decoy code to conceal malware strains such as CANFAIL and LONGSTREAM. Google also referenced a Russian campaign known as “Overload,” where AI voice cloning technology was allegedly used to imitate journalists in fabricated videos spreading anti-Ukraine narratives.

The report further examined the Android malware PromptSpy, previously documented by ESET, for its integration with Gemini APIs to automate interactions on infected devices.

Investigators identified an autonomous component called "GeminiAutomationAgent," which reportedly relies on a hardcoded prompt to help the malware evade AI safety mechanisms. Researchers explained that the prompt assigns the malware a harmless persona, enabling it to calculate interface geometry and interact with device functions more effectively.

Google researchers also warned that the malware appears capable of replaying authentication methods, including PINs and lock patterns, using AI-assisted techniques.

The company concluded that cybercriminals are increasingly scaling access to premium AI services through methods such as automated account generation, proxy relay systems, and shared account infrastructures.
Read more »
WhatsApp Payments
Bengaluru Startup Delayed Payments Fintech Solution Technology WhatsApp Payments

WhatsApp-Based Bengaluru Start-up Aims to Reduce Delayed Payment Woes

 

Delayed payments are a quiet but serious problem for small businesses, freelancers, tutors, and service providers, because the work may be complete while the money still remains stuck in follow-up cycles. In Bengaluru, a start-up called Lenda is trying to address that friction with a WhatsApp-first tool that automates reminders, supports negotiation, and helps users recover dues without creating awkward back-and-forth. 

The issue is not only financial but also practical, since chasing payments consumes time and can damage relationships between clients and providers. Many people already rely on WhatsApp for everyday communication, so the start-up is using that familiarity to make payment collection feel less like a formal recovery process and more like a normal conversation. 

Lenda’s approach is built around interactive messages instead of one-way reminders, which means a borrower can respond directly inside WhatsApp. The system lets recipients confirm payment, ask for extra time, raise a dispute, or even make a partial payment, which makes the process more flexible than a standard SMS reminder. That interaction matters because delayed payments often happen not just from unwillingness to pay, but also from timing problems, confusion, or simple forgetfulness. 

The start-up also tries to solve a structural problem for small operators such as teachers, class coordinators, and freelancers who collect money from many people at once. Its batch-reminder feature allows users to organize groups and send collective follow-ups, which reduces repetitive manual work and makes collections easier to manage. Lenda also includes late-fee options and a repayment score, aiming to encourage timely payment while giving businesses more control over overdue accounts. 

What makes the issue important is that delayed payments can disrupt cash flow, especially for small businesses that depend on regular incoming money to pay expenses and plan operations. By offering a “no-app” solution inside WhatsApp, Lenda is betting that the biggest barrier is not a lack of reminders, but the inconvenience and discomfort of asking for money repeatedly. That is why this Bengaluru start-up’s idea is less about messaging and more about fixing a common payment problem in a simpler, more human way.
Read more »
Secure Messaging
Apple Privacy Apple Security Encrypted RCS End To End Encryption iOS 26.5 iPhone Security Mobile Cyber Security RCS Protocol Secure Messaging

iOS 26.5 Introduces Private RCS Messaging and Core Feature Improvements


 

By introducing end-to-end encrypted RCS messaging between iPhone and Android devices for the first time, Apple has taken another step towards unifying secure cross-platform communication. 

In the update, Apple's messaging architecture has been significantly altered, extending advanced encryption protections beyond its proprietary ecosystem and into carriers' Rich Communication Services networks. This feature is currently being tested across major US networks and enables encrypted message exchange through the most recent version of Google Messages for Android, as well as Apple's native messaging experience, which is enhanced with visual encryption indicators and automatic activation mechanisms. 

RCS encrypted messages are currently available through a phased beta rollout to iPhone users running iOS 26.5 across supported carrier networks. Android compatibility is dependent on the latest version of Google Messages. It has been confirmed that encryption will be activated by default and gradually extended to both newly initiated and existing RCS conversations, eliminating the need for users to configure encryption manually.

Supported chats are now equipped with a dedicated lock icon that acts as a real-time confirmation layer, making sure messages are not readable while in transit between devices. Apple reiterated its commitment to privacy as its first priority, stating that iMessage remains fully encrypted within its native ecosystem, while the expansion of encrypted RCS provides an additional layer of security for cross-platform communication. 

According to industry analysts, the move is more of a strategic extension of Apple's broader device security framework than simply a messaging upgrade. According to Faisal Kawoosa, Founder and Chief Analyst at Techarc, the latest update enhances security assurances for Apple users outside of the iOS ecosystem, despite the fact that third-party messaging platforms will continue to be relevant.

With iOS 26.5, multiple system-level vulnerabilities are addressed, including issues relating to malicious media files and crafted text messages, causing application crashes, interface freezing, and potential denial-of-service exploitation scenarios before. 

Along with messaging overhaul, iOS 26.5 incorporates stability and security fixes. Modernizing the functionality of RCS itself, the update also brings advanced messaging capabilities, including high-resolution media transfer, typing indicators, read acknowledgement, reactions, and collaborative group chats across multiple devices. 

 Additionally, iOS 26.5 introduces a series of ecosystem refinements for personalization, subscription flexibility, and contextual user experiences in addition to its security-focused messaging upgrades. Apple has released an animated vertical light band wallpaper collection entitled Pride Luminance in honor of Pride Month, which shifts subtly as the device is unlocked, highlighting the importance of awareness of Pride Month. 

Apple continues to integrate adaptive visual design into iOS with its newest features, allowing users to customize wallpaper based on 11 predefined colour combinations or to create their own palette configurations. In addition to expanding subscription controls in the App Store, developers may also now offer monthly payment structures for discounted annual plans, a move that is intended to reduce upfront costs for long-term subscriptions while maintaining yearly commitments. 

The revised billing framework will require users who subscribe to annual packages through monthly payments to complete the payment cycle, regardless of whether the subscription is cancelled prior to the expiration date. Along with these additions, Apple has been continuing to expand its RCS rollout. Even though Rich Communication Services support was introduced with iOS 18 in 2024, it did not initially offer end-to-end encryption support, despite offering advanced messaging features such as high-resolution media sharing, typing indicators, read receipts, and advanced group chat features. 

In response to the integration of E2EE standards in the RCS specification by the GSMA last year, Apple has begun testing encrypted RCS support through the iOS 26 beta cycle and is preparing for a wider stable rollout. The availability of RCS support on iPhones continues to vary according to the network provider, because RCS functionality remains dependent on carrier-level implementation. 

Through the Messages settings panel, eligible users can manage the feature, displaying dedicated visual verification indicators, such as lock icons and encrypted session labels, in encrypted RCS chats. Aside from the refinement of core applications within Apple's release cycle, other core applications are being refined as well, including Maps updates that incorporate recommendations based on nearby trends and recent search behaviour, demonstrating the company's growing emphasis on contextually relevant software. 

Apple's iOS 26.5 not only extends feature parity between platforms but also reinforces its broader strategy to embed privacy and resilience deeper into everyday digital communication. By implementing end-to-end encryption for RCS conversations and simultaneously addressing media-handling vulnerabilities at the system level, the company is strengthening security controls around one of the most widely targeted layers of the mobile ecosystem. 

It reflects the growing industry trend towards interoperable, yet encrypted communication standards, where usability enhancements will increasingly coexist with enterprise-grade security protections and real-time threat mitigation.
Read more »
Phishing
2FA Cyber Attacks MFA Microsoft password Phishing

Microsoft Warns Passwords and SMS-Based 2FA Are No Longer Enough Against Modern Cyberattacks






Microsoft is intensifying its push toward passwordless security, warning that traditional passwords and older forms of two-factor authentication are becoming increasingly ineffective against modern phishing attacks powered by artificial intelligence.

In a statement released during World Passkey Day, Microsoft said the cybersecurity industry must reduce dependence on passwords and other “phishable” login methods by accelerating the adoption of passkeys. 

For years, technology companies encouraged users to strengthen account security by enabling two-factor authentication (2FA) or multi-factor authentication (MFA). Microsoft itself previously stated that MFA could block more than 99% of password-based attacks. However, cybercriminals have steadily adapted their tactics, particularly targeting SMS-based authentication systems through phishing pages, SIM-swapping schemes, session hijacking, and social engineering attacks.

The company now argues that passwords, even when paired with weak MFA methods like text-message verification codes, continue to leave accounts vulnerable. Microsoft described these older protections as “legacy” authentication methods that can still become entry points for attackers. 

Instead, Microsoft is promoting passkeys, which rely on cryptographic authentication rather than memorized passwords. A passkey stores a private digital key directly on a user’s device and only works on the legitimate website or application where it was created. Access is then confirmed through biometric verification, such as fingerprints or facial recognition, or through a device PIN. 

Security experts say this approach makes phishing significantly harder because passkeys cannot be reused on fake websites designed to imitate legitimate login pages. Unlike passwords or SMS codes, the authentication process is tied directly to the original domain. 

Microsoft also stressed that enabling passkeys alone is not enough if passwords and fallback authentication methods remain active on accounts. According to the company, weak backup options can still be exploited even after stronger protections are introduced. Microsoft has therefore continued removing older authentication systems across its ecosystem, including plans to eliminate security questions from password reset flows beginning in 2027. 

The urgency surrounding this transition has increased alongside the rapid growth of AI-generated phishing campaigns. Microsoft cited internal findings showing that AI-assisted phishing operations can achieve click-through rates as high as 54%, meaning more than half of targeted users may interact with malicious messages. 

Industry-wide adoption of passkeys is also accelerating. The FIDO Alliance estimates that more than five billion passkeys are already in use globally. Microsoft said hundreds of millions of users now sign into services such as OneDrive, Xbox, and Copilot using passkeys every day. 

Internally, Microsoft claims that over 99% of users within its environment now have access to phishing-resistant authentication methods. The company added that account recovery systems remain a critical security challenge because attackers increasingly target recovery processes instead of direct logins. 

Researchers and government agencies are broadly supporting the move toward passwordless security. The United Kingdom’s National Cyber Security Centre recently encouraged organizations and consumers to adopt passkeys, citing growing risks from AI-driven phishing and phishing-as-a-service platforms. 

Still, cybersecurity researchers caution that passkeys are not completely immune to attack. Recent academic research examining FIDO2 authentication methods found that while passkeys substantially raise the difficulty for attackers, sophisticated compromise techniques involving infected devices, session theft, or manipulated browser environments may still pose risks under certain conditions. 

Microsoft maintains that removing passwords and other phishable credentials remains essential as AI systems increasingly act on behalf of users across enterprise environments. If a single digital identity is compromised, attackers could potentially exploit connected AI agents to access systems, trigger workflows, and operate with existing permissions at machine speed. 

Read more »
Technology
AI Agent AI industry Artificial Intelligence Hermes openclaw python Technology

Hermes Agent Emerges as a Strong Challenger to OpenClaw in the Self-Learning AI Assistant Space

 



Artificial intelligence tools are increasingly allowing non-technical users to build software and automate tasks that previously required programming knowledge, and a new open-source AI agent called Hermes is becoming a major example of that shift.

The discussion gained momentum this week after reports circulated about a 78-year-old marketing executive with no coding background successfully creating a robotics application using only natural-language instructions. The application was reportedly built through the Reachy Mini ecosystem developed by Hugging Face, whose robot app marketplace has surpassed 300 live applications and approximately 10,000 deployed robots worldwide.

According to the shared account, the individual did not use Python programming or specialized robotics software during development. Supporters of AI-assisted development tools pointed to the example as evidence that conversational AI systems are reducing technical barriers that traditionally slowed software creation.

The development also reflects a broader trend across the AI industry. Newer AI agents are increasingly designed to retain information from previous interactions, improve their own workflows, and adapt to user behavior over time. Earlier this week, Anthropic introduced a feature called “Dreaming,” which allows AI agents to process earlier sessions in the background and generate new memory structures automatically. Meanwhile, Hermes Agent from Nous Research is pursuing a similar idea through persistent task learning and automated skill generation.

Hermes Agent, first released in February 2026, has quickly gained traction within the open-source AI community. The project reportedly has more than 135,000 GitHub stars and is distributed under the MIT license. It also includes over 40 built-in skills, which function as reusable instruction modules that help the system repeat previously learned workflows more efficiently.

One of Hermes’ defining features is its self-improving learning architecture. After completing a difficult or multi-step task, the agent enters what developers call a “Reflective Phase.” During this process, the system reviews its own actions, identifies successful execution patterns, and converts those patterns into reusable skill files. When a related task appears later, Hermes can retrieve the previously learned solution instead of generating a new workflow from the beginning.

The platform also uses a layered memory structure consisting of temporary session memory, long-term episodic memory stored through SQLite databases, and procedural memory tied to learned skills. Developers say the software can operate on low-cost virtual private servers, large GPU clusters, or serverless cloud environments. Hermes is also model-agnostic, allowing users to connect the framework to providers such as OpenAI, Anthropic, OpenRouter, Kimi, MiniMax, GLM, Nous Portal, or privately hosted AI endpoints.

Users can access the agent through Telegram, Discord, Slack, WhatsApp, Signal, email services, or command-line interfaces. The project’s latest update, v0.13.0, internally referred to as “The Tenacity Release,” reportedly introduced Google Chat integration as its twentieth supported platform. The update also added durable multi-agent coordination tools, automatic task recovery systems, retry budgeting controls, hallucination filtering mechanisms, persistent goal tracking for long-running tasks, automatic linting after file edits, and session recovery after unexpected gateway interruptions.

According to project details shared by contributors, the release included 864 code commits from 295 contributors in a single week and resolved eight critical security issues. One patched vulnerability reportedly involved a Discord-related flaw that could allow bots to message users across servers outside their intended access scope.

The installation process has also been simplified significantly. Hermes now uses a one-line curl installer that automatically configures dependencies such as Python 3.11, Node.js, ripgrep, and ffmpeg. During setup, the software can automatically detect existing OpenClaw environments and offer to import prior settings, memories, skills, and API credentials.

The growing comparison between Hermes and OpenClaw highlights a design shift occurring within the AI assistant ecosystem. OpenClaw originally gained attention by focusing heavily on messaging integrations and centralized orchestration across communication platforms. Hermes, by contrast, places continuous learning and automated self-improvement at the center of its architecture.

In practical terms, OpenClaw skills are generally predefined instruction sets written manually by users or generated beforehand through prompting. Hermes instead attempts to build those reusable workflows automatically by analyzing completed tasks after roughly every 15 tool interactions or after especially complex operations. Supporters argue this creates a compounding learning effect where the agent gradually improves with repeated use.

Despite the growing interest around Hermes, some developers caution against viewing it as a complete replacement for OpenClaw. OpenClaw still supports more than 24 messaging integrations, offers greater transparency through inspectable file-based memory systems, and has undergone broader public security review. Community discussions suggest that many advanced users currently operate both systems together, using OpenClaw for orchestration while relying on Hermes for adaptive learning capabilities.

Researchers tracking the rapid development of AI agents believe these systems are moving beyond traditional chatbot behavior and evolving into persistent digital assistants capable of handling long-running, multi-step workflows. However, cybersecurity analysts also warn that systems with autonomous memory creation and broad platform access may introduce additional security and privacy risks if governance and safeguards fail to evolve alongside the technology.

Read more »
Malware Attack
Cyber Attacks Cyber Breaches Cyber websites data security Fake Windows Linux Linux Malware Malware Attack

JDownloader Website Breach Spreads Malware Through Fake Windows and Linux Installers

 

In early May 2026, the official website for JDownloader was compromised, causing users to unknowingly download infected installers instead of legitimate software. During the two-day breach window, attackers replaced Windows and Linux setup files with malicious versions carrying hidden malware. Researchers later discovered that the Windows payload deployed a stealthy Python-based remote access trojan capable of giving attackers control over infected systems. 

Because the files appeared authentic and came directly from a trusted source, many users installed them without suspicion. JDownloader remains one of the most widely used download automation tools, supporting downloads from hosting services, streaming sites, and premium file-sharing platforms across Windows, Linux, and macOS. Its long-standing reputation and large user base made the attack especially dangerous, as users naturally trusted downloads from the official website. 

The issue first gained attention after a Reddit user reported Microsoft Defender warnings while downloading updated installers from the JDownloader website. The files showed suspicious digital signatures linked to unknown names like “Zipline LLC” and “The Water Team” instead of AppWork GmbH, the legitimate developer. Community concern quickly spread online, prompting the development team to investigate. 

Soon after, JDownloader confirmed that attackers had exploited an unpatched flaw in the site’s content management system to modify download links and redirect users toward malicious third-party installers. Developers stated that the compromise was limited to public-facing web content and did not extend to deeper server infrastructure or operating system-level access. The team later clarified that only the Windows “Alternative Installer” downloads and Linux shell installer links were affected. 

Other distribution channels, including macOS packages, Flatpak, Winget, Snap releases, in-app updates, and the main JAR package, remained secure throughout the incident. Developers urged users to verify installer authenticity by checking digital signatures within file properties. Legitimate files should display a verified signature from AppWork GmbH, while unsigned installers or files signed by unfamiliar publishers should be avoided immediately. 

Cybersecurity researcher Thomas Klemenc later analyzed the malicious Windows files and found they acted as loaders for a heavily obfuscated Python-based remote access tool. According to his findings, the malware could execute remote commands through command-and-control servers, silently turning infected devices into attacker-controlled systems. Analysis of the Linux shell installer also uncovered injected malicious code designed to download disguised payloads from suspicious domains. 

Once executed, the malware installed hidden binaries, created persistence mechanisms, elevated privileges using root-level configurations, and disguised itself as legitimate Linux system processes to avoid detection. Experts noted that parts of the Linux malware remain difficult to fully understand because the payload was heavily protected using obfuscation tools like Pyarmor, limiting deeper analysis. 

Although JDownloader stressed that only users who downloaded and executed installers during the breach window were at risk, security professionals strongly recommend reinstalling operating systems on infected machines. Since arbitrary code execution was possible, experts also advise resetting all passwords after cleaning affected devices due to potential credential theft. 

The attack reflects a growing cybersecurity trend in which hackers target trusted software platforms to distribute malware through compromised downloads. Similar incidents recently affected CPU-Z, HWMonitor, and DAEMON Tools, where attackers replaced legitimate installers with infected versions carrying hidden malware.  

As supply chain attacks continue increasing, cybersecurity experts stress the importance of checking digital signatures carefully and avoiding suspicious downloads, even on trusted software platforms.
Read more »
WolfSSL flaw
AI cybersecurity Anthropic Claude Mythos Preview Project Glasswing software vulnerabilities Vulnerabilities and Exploits WolfSSL flaw

Anthropic’s Project Glasswing Detects Over 10,000 Critical Software Vulnerabilities Worldwide

 

iArtificial intelligence company Anthropic has revealed that its cybersecurity initiative, Project Glasswing, has successfully identified more than 10,000 high- and critical-severity vulnerabilities across globally significant software systems since the program was introduced last month.

The initiative was designed as a defensive cybersecurity program aimed at strengthening critical software infrastructure worldwide. Through Project Glasswing, around 50 trusted partners receive early access to Claude Mythos Preview — an advanced AI model capable of autonomously discovering vulnerabilities in widely used software before malicious actors can exploit them.

According to Anthropic, 6,202 of the detected vulnerabilities were categorized as high or critical severity and affected over 1,000 open-source projects. Further review confirmed 1,726 of these findings as legitimate true positives, while 1,094 vulnerabilities were assessed as either high or critical in severity.

Among the major discoveries was a critical security flaw in WolfSSL identified as CVE-2026-5194, carrying a CVSS score of 9.1. The vulnerability could potentially allow attackers to forge certificates and impersonate legitimate services. Anthropic noted that the initiative has already contributed to 97 vulnerabilities being patched upstream along with the release of 88 security advisories.

"The relative ease of finding vulnerabilities compared with the difficulty of fixing them amounts to a major challenge for cybersecurity," Anthropic acknowledged. "Confronting this challenge successfully will make our software far safer than before."

The announcement comes amid a broader rise in AI-assisted vulnerability discovery, with software vendors releasing patches at an unprecedented pace. Microsoft recently indicated that the number of monthly security patches is expected to continue increasing over time.

Cybersecurity firm XBOW described Mythos Preview as "a major advance" that is "substantially better than prior models at finding vulnerability candidates" and "adept at analyzing source code with a security mindset." Researchers have also observed the model’s effectiveness in converting vulnerabilities into complete end-to-end attack chains.

Anthropic highlighted that the capabilities of Mythos Preview extend beyond vulnerability detection. In one reported incident, a banking partner participating in Glasswing used the AI model to identify and block a fraudulent wire transfer worth $1.5 million after a threat actor compromised a customer’s email account and attempted spoofed phone calls.

The company warned that AI models with capabilities similar to Mythos could become widely accessible in the near future, prompting a need for organizations to accelerate their patch management processes. Oracle has already transitioned to a monthly patch cycle to respond more quickly to critical security vulnerabilities.

"Network defenders should shorten their patch testing and deployment timelines," Anthropic said. "These include steps like hardening networks' default configurations, enforcing multi-factor authentication, and keeping comprehensive logs for detection and response."

Anthropic also announced the launch of its Cyber Verification Program, which allows verified security researchers to use its AI models without standard guardrails for legitimate cybersecurity activities such as penetration testing, vulnerability research, and red teaming. The move mirrors OpenAI’s Daybreak initiative, which enables defenders to work with GPT-5.5-Cyber for specialized security workflows.

Despite their advanced capabilities, models such as Mythos Preview and GPT-5.5-Cyber have not yet been publicly released due to concerns surrounding potential misuse and the absence of sufficient safeguards against large-scale abuse.

"Glasswing helps the most systemically important cyber defenders gain an asymmetric advantage," it pointed out. "However, there is an urgent need for as many organizations as possible to shore up their cyber defenses. We hope that our generally available models, and the new tools, resources, and research we're providing to accompany them, will support those organizations to improve their cybersecurity posture."
Read more »
Vulnerabilities and Exploits
Claude AI Google Ads Mac Users User Security Vulnerabilities and Exploits

Hackers Abuse Google Ads and Claude.ai Chats to Spread Mac Malware

 

Cybercriminals are once again abusing trust, and this time they are combining Google Ads with Claude.ai shared chats to push malware onto Mac users. The campaign targets people searching for terms like “Claude mac download,” where sponsored results appear to point to the legitimate claude.ai domain but actually lead to malicious installation instructions. Security researcher Berk Albayrak first identified the scheme, and confirmed that attackers are using the tactic in active campaigns. 

The attack works because it looks believable at first glance. Users click a sponsored search result, land on a public Claude chat, and see what appears to be an official “Claude Code on Mac” guide, sometimes even attributed to Apple Support. That page then tells them to open Terminal and paste a command. Instead of installing useful software, the command quietly downloads and runs malware on the victim’s Mac.

What makes the operation especially dangerous is the way it blends legitimate services with deception. The ad itself can show the real claude.ai domain, which helps the link look safe, while the malicious instructions are hidden inside Claude’s shared chat feature. In some variants, the payload is linked to MacSync-style infostealer behavior, aimed at harvesting browser credentials, cookies, and Keychain data. Researchers also reported that multiple malicious chats were being used, showing that the operators are testing and rotating infrastructure. 

The campaign is a strong reminder that search results and AI platforms are not automatically trustworthy just because they appear familiar. Attackers increasingly rely on “clickfix” tactics, where the victim is convinced to copy and run a command manually, bypassing many traditional download warnings. That user action becomes the infection point, making the social engineering as important as the malware itself.

Mac users should avoid sponsored search results when looking for software downloads and instead go directly to the official site by typing the address themselves. Any chat, guide, or support page that instructs users to paste Terminal commands should be treated with caution, especially if it claims to come from Apple or a well-known AI service. The broader lesson is simple: when an instruction asks you to run code on your own computer, pause and verify before acting.
Read more »
ShinyHunters
Canvas LMS Data Breach Instructure Breach Login Page Defacement Ransomware Attack School Cyberattack ShinyHunters

Threat Campaign Targets School Login Systems After Alleged Instructure Hack


 

The initial appearance of a routine service disruption within one of the most widely used academic learning platforms in the world quickly evolved into a significant cybersecurity issue as threat actors associated with the ShinyHunters group allegedly compromised Instructure's Canvas system. 

A large number of educational institutions experienced widespread operational instability as a result of the incident, which exposed sensitive academic and identity-related records, disrupted coursework timelines, and resulted in the defacement of several school authentication portals. 

A growing concern over the potential release of a data set reportedly affecting thousands of institutions as well as hundreds of millions of students and employees led Instructure to reveal that it had reached an agreement with the unauthorised actor responsible for the intrusion language that cybersecurity analysts interpreted as an indication of ransom negotiations. ShinyHunters collective claims to have successfully compromised Instructure's infrastructure for the second time in just a few weeks, further escalating the issue. 

The breach resulted in school authentication portals were made public and were affected in addition to backend systems. The incidents took place during final examination periods across several institutions using Canvas, causing even more disruption for administrators, educators, and students experiencing intermittent outages as a result of the earlier intrusion disclosed on April 30.

The Instructure platform had acknowledged that "criminal threat actors" were responsible for unauthorized access to parts of its environment, but subsequent activity indicates the attackers were still able to manipulate externally accessible services. 

When threat actors were reportedly injected malicious HTML components into Canvas login pages, unauthorized message prompts were found attributed to ShinyHunters, effectively defacing the authentication screens utilized for coursework management, assignment submissions, and academic communication, multiple Canvas login pages were later found displaying unauthorized messages attributed to ShinyHunters.

According to the message posted by the group, the allegedly stolen data will be made public on May 12 unless the company enters into a "settlement" negotiations. Parts of Instructure's online infrastructure appeared unstable during the escalation process, with some services intermittently returning "too many requests" errors while Canvas displayed maintenance notices indicating ongoing remediation and containment efforts throughout the company's network infrastructure. 

According to further disclosures, the breach affected a wide spectrum of academic stakeholders, including students, faculty, and institutional staff, with portions of information reportedly relating to minors. Despite Instructure's claims that passwords and highly sensitive authentication credentials were not compromised, the attackers are said to have obtained substantial amounts of information regarding personal identification and platform usage, such as usernames, e-mail addresses, student identification numbers, and private communications exchanged within the learning management system. 

According to the company, the initial compromise was terminated, remediation measures were implemented across the affected systems, and Canvas services were restored after containment procedures were initiated to prevent additional intrusions. However, ShinyHunters later stated it had successfully breached the platform again, this time targeting institution-specific authentication portals, thereby putting the company under pressure to enter into a settlement negotiation related to the earlier data theft, despite these efforts. 

As part of the extortion attempt, the group used stolen data as a means of coercion following network intrusions, which is a well-established operational pattern, however, the apparent recurrence of unauthorized access raised concerns regarding residual vulnerability issues within Instructure's network infrastructure. Canvas was brought offline once again following the second disruption, prompted the company to remove the component identified as being at the root of the incident  the Free-for-Teacher environment. 

Instructure acknowledged in an updated incident disclosure that investigators had identified a vulnerability associated with support ticket functionality within the Free-for-Teacher system, which threat actors allegedly exploited to facilitate the latest security breach. By putting the incident on its leak portal, ShinyHunters had earlier accepted public responsibility for the initial intrusion. 

The tactic is commonly used by ransomware and extortion-focused groups to increase pressure on targets by threatening data release under controlled circumstances. In the wake of the recent compromise, the attackers have attempted to reach out directly to media outlets regarding the defaced Canvas login pages, suggesting they are attempting to escalate the attack not only against Instructure but also against the thousands of educational institutions that rely on the platform for their operations. During ongoing negotiations regarding the previously stolen data, cybersecurity analysts viewed the public defacement as an attempt to amplify reputational and operational pressures. 

In spite of the fact that there is no clear indication of how the school-specific authentication pages were compromised, ShinyHunters officials have indicated the breach has been a separate one from the original attack, but declined to provide any further technical information regarding the method used to gain access to the system. 

The group claims to have stolen data from nearly 9,000 educational institutions around the world; these records are believed to belong to approximately 231 million people. Following the earlier compromise, the group claimed to have exfiltrated information related to nearly 9,000 educational institutions. 

A key component of the campaign was a mirroring of the threat group's established operating model, which is typically composed of a combination of network intrusion, public exposure of victims through leak sites, and sustained extortion efforts to maximize financial leverage following the theft of large amounts of data. There has been an increased focus on security architecture of cloud-based education platforms in the wake of the incident, which has become a critical infrastructure for academic operations worldwide.

In addition to disrupting coursework and institutional systems for the immediate period, the exposure of student communications and identity-linked records, particularly involving minors, demonstrates the long-term risks associated with large-scale compromises of digitally centralized learning environments. 

During the remediation and forensic investigation efforts, Instructure is likely to establish the breach as a landmark in the field of ransomware and extortion, which increasingly target educational technology ecosystems where operational urgency and reputational pressure can lead to high-stakes cybersecurity incidents.
Read more »
data security
AI coding tools Consumer Data Exposure Corporate Data Breach cybersecurity vulnerabilities Data Breach data security

AI Coding Tools Expose Thousands of Apps With Sensitive Corporate Data Online

 

Thousands of web applications built using AI coding tools have been found publicly accessible online without proper security protections. Researchers at RedAccess identified more than 5,000 exposed apps tied to companies, many revealing private information to anyone with the correct URL. Employee records, customer conversations, system plans, and financial files were among the exposed materials. The problem wasn’t faulty code but missing security setup steps that many users overlooked. 

In many cases, public access remained enabled long after deployment, creating silent data leaks that went unnoticed for months. Many of the vulnerable apps were created using platforms like Replit, Netlify, Base44 owned by Wix, and Lovable. Nearly 2,000 apps appeared to contain genuine sensitive information, including advertising spending reports, company strategy documents, chatbot logs, customer contact details, hospital personnel records, and financial summaries. 

According to RedAccess researcher Dor Zvi, the issue is linked to the rise of “vibe coding,” where non-technical employees use AI tools to rapidly build and publish web applications. Since these platforms make development extremely simple, apps can go live within minutes without any review from engineering or cybersecurity teams. Researchers found the exposed apps through basic Google and Bing searches because many AI coding services host projects publicly on shared domains by default. 

Some applications exposed private information without requiring logins, while others reportedly allowed outsiders to gain administrative control over backend systems. The exposed data covered multiple industries. Hospital staff schedules listing doctors’ identities appeared alongside marketing strategy presentations, shipping records, retailer chatbot conversations, and detailed advertising campaign budgets. Such leaks could expose sensitive competitive information, including business planning timelines and financial allocations. 

The investigation also uncovered phishing websites hosted directly on AI coding platform domains. These fake pages impersonated major companies including Bank of America, Costco, FedEx, Trader Joe’s, and McDonald’s. The platforms disputed parts of the findings while acknowledging that publicly accessible apps existed. Amjad Masad said users choose whether apps remain public or private. Lovable emphasized that creators are responsible for configuring security correctly, while Wix stated weakening protections requires deliberate user actions. 

Security experts argue the broader issue remains serious because AI coding tools rarely enforce strong safeguards automatically. Many employees using them lack training in authentication systems or permission controls, allowing insecure deployments to slip through unnoticed. Researchers say the situation resembles earlier waves of exposed Amazon S3 cloud storage buckets, where confusing defaults and user mistakes left sensitive files publicly accessible. 

AI-powered coding platforms may now be accelerating similar risks on a larger scale as businesses increasingly rely on AI tools for internal dashboards, marketing systems, client portals, and reporting applications. Experts also warn the true scale may be far larger. The 5,000 discovered apps only included projects hosted directly on AI platform domains. Thousands more could exist on privately owned domains that standard searches cannot easily detect. 

As AI-generated development grows rapidly, companies are now under pressure to strengthen oversight, improve employee training, and introduce stricter security reviews. Without stronger safeguards, fast AI-assisted app creation could continue exposing confidential corporate and personal information online.
Read more »
User Privacy
Android Trojan Data Harvesting Data Leak malware User Privacy

Millions of Devices at Risk: New Trojan Monitors Smartphones

 

A menacing new Trojan has emerged that puts millions of smartphone devices worldwide at risk, according to recent cybersecurity reports. This sophisticated malware specifically targets Android devices and has already infected thousands of users across 143 countries. The Trojan's ability to monitor smartphones in real-time represents a significant evolution in mobile cyberthreats, with security researchers warning that the actual infection count could be far higher than currently detected.

The malware spreads primarily through seemingly legitimate websites that trick users into downloading malicious applications. Once installed, the Trojan grants hackers complete remote control over compromised devices, enabling live monitoring of user activities. Security firm Zimperium zLabs identified similar dangerous Trojans like Arsink, which impersonates popular brands including WhatsApp and TikTok to evade detection. The infected devices can have their audio recorded, text messages read, and even be wiped completely by attackers. 

This Trojan's most alarming capability is its live monitoring feature combined with coordinated attack systems. Beyond stealing credentials, the malware transmits live screen content to remote servers, creating a continuous visual feed that allows attackers to observe activity and intercept authentication steps in real time. Encrypted communication channels connect infected devices to centralized command systems that coordinate attacks and distribute updated instructions, managing thousands of compromised devices simultaneously. The infection has created a massive footprint, with Egypt reporting around 13,000 compromised phones, Indonesia approximately 7,000, and Iraq and Yemen each with 3,000 infections. 

The Trojan harvests an extensive range of sensitive data including SMS messages, call logs, contacts, device location, and Google account information. It can steal user accounts in messengers and social networks, stealthily send messages on behalf of victims, monitor browser activities, replace links, swap numbers during calls, and intercept SMS messages. Previous similar malware campaigns have already stolen at least $270,000 worth of cryptocurrency, suggesting the financial damage from this new Trojan could be substantial. 

Experts recommend several critical protection measures to safeguard against this threat. Users should only download applications from official app stores like Google Play, avoid clicking links from suspicious websites, and keep their Android operating system updated with the latest security patches. Google has warned that over 40% of Android devices remain vulnerable because they run outdated versions without security support. If your smartphone brand no longer provides security updates, experts strongly recommend considering a new device to protect your personal data.
Read more »
WhatsApp
Bugs Cloud Data Meta Technology WhatsApp

WhatsApp Fixed Two Security Bugs via It's Bug Bounty Program

WhatsApp Fixed Two Security Bugs via It's Bug Bounty Program

Meta recently released a security advisory in May revealing two bugs in WhatsApp were found through its bug bounty program. But these bugs were patched and were not exploited in the wild by the threat actors. Both bugs are now patched.

About two bugs

The first bug is tracked as CVE-2026-23863, a Windows specific problem. This bug was maliciously crafted with hidden “NUL BYTES” hidden within the filename, to trick WhatsApp into showing it as one filetype such as an authorized PDF while pretending to be running as an executable once opened. Meta fixed this patch in April on both platforms.

The second vulnerability, tracked as CVE-2026-23866 impacted both android and iOS users. The attack tactic involved partial authorization of AI rich response texts for Instagram Reels shared within Whatsapp. A threat actor could possible launch another user’s device to access media content through an arbitrary URL, such as launching OS level custom URL scheme handles. This flaw was patched in April on both platforms.

Severity

The two bugs were given medium severity by researchers. WhatsApp has verified that no bug was abused.

Both were rated medium severity, and WhatsApp confirmed there's no evidence either was actually abused.

The impact

These kind of reporting get sidelined by glossy and infamous threat. For instance the recent SMS pumpoing attacks increasing phone bills, or phishing campaigns that used messaging apps as entry points, and lastly the attack on educational institutes that compromised Canvas and Instructure, leaking hundreds of GBs of data.

But Whatsapp did a good job in finding and fixing the flaw before cybercriminals could exploit them and cause harm. The bug bounty program of WhatsApp has been going on for fifteen yesr, and the recent patches show it it is still reliable.

What should users do?

Simple advice: always keep your phones and app updated. 

There has never been a better moment to use secure communications services like WhatsApp or Signal. The truth is that Meta does a great job of keeping the app and its users safe and secure, despite some security concerns of its own, such as the recently reported phishing attempts using the encrypted messenger as part of the exploit chain and a spyware threat targeting iOS users.

Read more »
Subscribe to: Posts (Atom)