Search This Blog
Powered by Blogger.

Blog Archive

Labels

Footer About

Footer About

Labels

Load More
Trendy Right Now

Popular Posts

Showing posts with label software. Show all posts
technology
ai agents apis business data consulting MCP Salesforce software technology

Salesforce’s Headless 360 Pushes Enterprise Software Beyond the Browser

 




Salesforce is preparing for a future in which employees may no longer need to open Salesforce to use it.

At TDX 2026, CEO Marc Benioff described the shift with the line, “Our API is the UI,” as the company introduced Headless 360. The platform makes Salesforce capabilities, including Customer 360, Agentforce and Slack, accessible through APIs, Model Context Protocol (MCP) tools and command-line interfaces (CLI), allowing applications and AI agents to interact with Salesforce without relying on its traditional browser interface. Salesforce says its Headless 360 MCP server can support operations including querying and updating records, managing permissions, working with Apex and interacting with platform events.

The change challenges a model Salesforce spent decades building: software operated primarily by humans through screens and sold largely through user-based licensing.

If an AI agent performs the work, the traditional per-seat model becomes harder to justify. An agent does not need a dashboard or training programme in the same way an employee does. It needs authenticated access to data, tools and workflows.

Salesforce is already experimenting with consumption-based pricing. Its Agentforce model includes Flex Credits, which customers can use for agent actions, alongside conversation-based and user-based pricing. Salesforce lists 100,000 Flex Credits at $500, while certain Agentforce services can also be priced according to successful outcomes.

That transition could also affect the Salesforce consulting ecosystem. Implementation work historically centred on configuring screens, workflows and processes for employees. As agents take over more workflows, organizations may instead spend more on data quality, permissions, API architecture, agent governance and testing.

Salesforce has a reason to disrupt itself before competitors do.

AI-native platforms can be designed around APIs and autonomous agents without inheriting the assumptions of traditional enterprise software. By opening Salesforce to agents, the company is betting that its strongest asset is not the interface but the business data, permissions and workflows underneath it.

That makes governance a central part of the strategy.

Salesforce's Einstein Trust Layer is designed to keep Agentforce grounded in enterprise data while respecting existing access controls. Salesforce describes capabilities including dynamic grounding, secure data retrieval, auditability and zero-data-retention arrangements with external model providers.

But making Salesforce accessible through MCP and external AI systems creates another risk: the company no longer fully controls the interface through which users interact with its platform.

A sales manager could eventually ask an external AI agent to analyse pipeline data, update opportunities, trigger Salesforce workflows and coordinate information across Slack, Salesforce and other enterprise systems. The AI layer becomes the operating interface while Salesforce functions as the underlying system of record.

MCP also introduces new security considerations. Research has identified threats including tool poisoning and prompt injection, where malicious instructions embedded in tools or outputs can influence an agent's behaviour. The U.S. National Security Agency has similarly warned about cascading prompt-injection risks in MCP environments, where one agent's output can become another system's input.

The pricing problem remains unresolved as well. Agent actions vary enormously in complexity. Updating a contact record is not equivalent to autonomously completing a sales renewal, making a simple “pay per action” model difficult to align with business value.

Salesforce's Headless 360 strategy therefore represents more than a move away from browsers. It is a test of what enterprise software is worth when humans are no longer its primary operators.

Interfaces can be replaced. What is harder to replace is trusted business data, permission architecture, proprietary workflows and the infrastructure required to let autonomous systems act safely.

Salesforce is betting that those foundations will remain valuable.

The risk is that by making them accessible to external agents, it could also help those agents become the new interface between enterprises and Salesforce itself.

Read more »
Supply Chain Attacks
Amazon malware Sapphire Sleet software Supply Chain Attacks

Amazon Attributes Earlier npm Supply Chain Attacks to North Korea's Sapphire Sleet

 



Amazon has linked a series of high-profile npm supply chain compromises spanning 2025 and 2026 to the North Korean threat group Sapphire Sleet, suggesting that attacks initially viewed as isolated incidents may instead represent a coordinated campaign targeting widely trusted open source software.

In a threat intelligence report published on July 29, Amazon assessed with medium confidence that the same actor responsible for the March 2026 compromise of the popular JavaScript package axios was also behind earlier attacks involving the npm packages debug and chalk, as well as a lesser-known package called typo-crypto. The assessment expands the scope of what security researchers now believe to be a sustained operation aimed at infiltrating software supply chains through compromised maintainer accounts.

The September 2025 incident involving debug and chalk drew widespread attention after attackers successfully phished an npm maintainer using a fraudulent npm website designed to harvest credentials. Once access was obtained, malicious updates were published to multiple packages collectively downloaded billions of times each week. Rather than infecting developers' systems directly, the malicious code targeted cryptocurrency users by intercepting browser-based wallet activity and replacing legitimate transaction addresses before users approved transfers.

At the time, security firms including Aikido Security and Wiz documented the compromise and analyzed its technical behavior, but neither publicly attributed the operation to a specific threat actor. Amazon's latest research represents the first detailed effort to connect that incident with a broader campaign linked to North Korea.

According to Amazon, investigators uncovered additional evidence while examining the March 2026 axios compromise. During that investigation, analysts identified a domain registered in 2025 that ultimately led them to a previously overlooked npm package named typo-crypto. Although the package attracted relatively few downloads, Amazon believes it served as an early testing ground for techniques that later appeared in attacks targeting far more widely used libraries.

The company argues that the campaigns share several operational characteristics, including the deployment of trojanized packages, overlapping command-and-control infrastructure, similarities in malicious code, and the use of social engineering to gain access to trusted maintainer accounts before distributing compromised package updates. Based on those shared indicators, Amazon believes the incidents form part of the same long-running operation.

However, the report has also prompted discussion within the security community regarding the strength of the evidence supporting the attribution. While Amazon outlines common tactics and infrastructure across the campaigns, the report does not publicly specify which individual indicators directly connect each incident. As a result, some researchers have noted that although the overall assessment appears plausible, additional technical evidence would help strengthen the case for linking every campaign to the same actor.

The technical methods employed across the attacks also differed substantially.

The malicious code inserted into debug and chalk functioned primarily within web browsers. It intercepted browser APIs associated with cryptocurrency wallets and modified transaction destinations before users authorized transfers. Security researchers observed that the attack did not rely on npm lifecycle scripts or establish persistent malware on infected systems.

By contrast, the March 2026 axios compromise involved a post-install payload that executed during package installation, allowing attackers to deploy additional malicious components. Amazon also identified similarities between that campaign and the earlier typo-crypto package, which contained a disguised file named core.js. The file reportedly activated only after receiving a specific trigger and then retrieved an operating system-specific second-stage payload from a remote command-and-control server.

Amazon identified infrastructure associated with the malicious package, including the domain npmjs.store and the IP address 216.74.123.126. The company also noted that the malware concealed portions of its functionality using Base64 encoding combined with an XOR-based obfuscation routine.

Further examination of the npm registry revealed additional irregularities surrounding typo-crypto. The package appeared to have been published only once, with no earlier legitimate versions preceding the malicious release. Its metadata closely resembled that of the legitimate crypto-js project, including copied descriptions and keywords, while advertising a version number ahead of crypto-js itself. Those characteristics suggest the package was created from the outset to impersonate an established library rather than resulting from the compromise of an existing maintainer account.

Amazon also referenced the Open Source Vulnerabilities database entry MAL-2026-3400 in connection with typo-crypto. Registry records indicate that the package remained publicly available at the time researchers reviewed it. Although it did not declare an install script capable of automatically executing malicious code upon installation, investigators confirmed that the embedded core.js file contained trigger values consistent with Amazon's analysis. Researchers also identified discrepancies involving one published SHA-256 hash, leaving open the possibility of either a documentation error or a hash corresponding to a different sample.

The attribution aligns with assessments previously made by other major cybersecurity vendors regarding the axios compromise. Google attributed that incident to the cluster it tracks as UNC1069, citing malware known as WAVESHAPER.V2 together with infrastructure previously associated with the group. Microsoft separately attributed the operation to Sapphire Sleet, a financially motivated North Korean threat actor also tracked under several alternative names by different security vendors.

Threat intelligence researchers generally consider medium-confidence assessments to indicate that multiple independent indicators support an attribution while acknowledging that additional evidence could alter future conclusions. In this case, Amazon's analysis represents another step toward understanding the relationship between several supply chain attacks, even as researchers continue examining the technical links connecting them.

Open source software ecosystems remain attractive targets because compromising a single trusted package can affect thousands of downstream applications and organizations. Libraries such as debug, chalk, and axios are deeply embedded throughout the JavaScript ecosystem, meaning malicious updates have the potential to propagate rapidly across development environments before they are detected.

The incidents have also renewed attention on software supply chain security. Earlier this month, npm introduced version 12, disabling dependency lifecycle scripts by default to reduce opportunities for post-install malware execution. The registry has also begun scanning newly published packages for malicious code before they become available to users. While these measures help address certain attack techniques, security researchers caution that they do not eliminate the risk posed by compromised maintainer accounts obtained through phishing or other forms of social engineering.

As open source ecosystems continue to expand, security experts expect attackers to increasingly focus on trusted maintainers rather than exploiting software vulnerabilities alone. The latest attribution from Amazon underlines the growing role of identity-based attacks in software supply chain operations and emphasises the continuing need for stronger maintainer protections alongside technical safeguards.

Read more »
VPNs
compromised credentials Cyber Security Emails MFA Phishing ransomware software Sophos VPNs

Phishing and Compromised Identities Replace Software Exploits as Leading Ransomware Entry Ooint, Sophos Reports





Phishing campaigns, malicious emails and compromised credentials have overtaken software vulnerability exploitation as the leading entry points for ransomware attacks, according to Sophos' State of Ransomware 2026 report, signalling that threat actors are placing greater focus on stealing identities than breaking into unpatched systems.

The report is based on responses from 2,158 IT and cybersecurity leaders across 17 countries whose organisations experienced ransomware attacks during the previous year. While ransomware groups continue to encrypt data in a large share of incidents, the findings point to a clear change in how attackers gain their initial foothold inside enterprise networks.

Malicious email accounted for 26% of ransomware attacks, making it the most common root cause identified by respondents. Phishing followed closely at 24%, while compromised credentials were responsible for another 23% of incidents. In comparison, exploited vulnerabilities accounted for 18% of attacks, a sharp decline from 32% reported in the previous edition of the survey.

Taken together, email-based attacks and stolen credentials were responsible for nearly three-quarters of reported ransomware intrusions, showing that attackers are increasingly relying on social engineering and identity compromise instead of searching for vulnerable internet-facing systems.

Sophos also found that 67% of organisations described the ransomware incident as the most serious identity-related attack they encountered during the past 12 months. Across the surveyed organisations, almost four out of five ransomware attacks originated through compromised identities, placing user accounts and authentication systems at the centre of modern ransomware operations.

The findings also challenge a common assumption about multifactor authentication. Sophos reported that MFA had already been deployed in 97% of attacks where compromised credentials were identified as the root cause, yet attackers still succeeded in gaining access.

According to the company, the figures do not mean MFA has become ineffective. Instead, they point to weaknesses in deployment and the growing sophistication of credential theft techniques. Some organisations may have protected only part of their infrastructure, leaving legacy systems, remote access services or administrative interfaces outside MFA coverage. Those gaps can provide attackers with alternative routes into corporate environments.

Attackers have also refined methods for bypassing authentication protections. Adversary-in-the-middle phishing kits can intercept authentication sessions, while stolen browser cookies and authenticated session tokens allow attackers to access accounts without repeatedly triggering MFA challenges. MFA fatigue attacks, where users are bombarded with repeated authentication requests until one is approved, continue to be used against organisations relying on push-based authentication.

Among organisations using MFA, one-time passwords, push notification applications and passkeys were the most widely deployed authentication methods. FIDO2 security keys ranked behind those options despite offering one of the strongest defences against phishing because authentication is tied to legitimate websites and cannot be replayed through fake login pages.

Although software vulnerabilities no longer ranked as the leading ransomware entry point, Sophos cautioned that organisations should not reduce their focus on patch management. Instead, the report recommends pairing vulnerability remediation with stronger identity security controls to reduce opportunities for attackers to obtain valid credentials.

To reduce the risk of email-driven intrusions, Sophos recommends deploying advanced email filtering alongside domain authentication technologies including DMARC, DKIM and SPF. These controls help organisations verify legitimate senders, detect spoofed domains and prevent fraudulent emails from reaching employees. The company also recommends regular phishing awareness training to help users identify increasingly convincing social engineering campaigns.

Beyond email security, Sophos advises organisations to strengthen identity protection through Identity Threat Detection and Response (ITDR), enforce MFA across every access point and routinely review both human and machine identities to remove unnecessary privileges, dormant accounts and outdated credentials that could be abused during an attack.

Chet Wisniewski, director and global field chief information security officer at Sophos, said organisations with stronger ransomware resilience typically rely on multiple defensive layers rather than a single security control. Network segmentation can slow lateral movement after an initial breach, Zero Trust Network Access (ZTNA) reduces dependence on traditional VPNs, and continuous threat detection gives security teams more opportunities to identify malicious activity before ransomware spreads across the network.

The report also found that ransomware operators successfully encrypted data in 56% of reported attacks. Although median ransom demands and payments have fallen compared with previous years, the findings show that attackers continue to achieve their primary objective once they obtain access. For defenders, protecting identities has become just as important as patching software, with user accounts, credentials and authentication systems now representing the most frequently targeted path into enterprise environments. 

Read more »
software
Cybersecurity Hyadina Microsoft PoisonX RaaS ransomware software

Microsoft-Signed Driver Used to Disable Security Software in GodDamn Ransomware Attacks

 


A ransomware group known as Hyadina has been observed using a Microsoft-signed Windows kernel driver to disable endpoint security software before deploying its latest ransomware variant, GodDamn, according to researchers at Symantec. The campaign combines trusted administrative software, publicly available offensive security tools and a signed kernel driver to establish control over victim environments before encrypting systems.

Hyadina has operated as a ransomware-as-a-service (RaaS) group for approximately four years, evolving its malware from earlier variants known as Beast and Monster to its current GodDamn locker. The group primarily targets organizations in the United States while reportedly avoiding victims in former Soviet countries. Previous attacks have affected organizations across healthcare, manufacturing, education and several other industries.

Symantec was unable to determine how the attackers initially gained access to the compromised environment. The first confirmed activity appeared on May 29, when an unauthorized copy of AnyDesk was discovered inside the Music folder of an infected system. Although AnyDesk is legitimate remote access software widely used for IT support, threat actors frequently abuse remote monitoring and management applications because they provide persistent access while blending into normal administrative activity.

The following day, the attackers deployed an executable named symantec.exe, which installed a kernel-mode driver called PoisonX. Running in the Windows kernel gives a driver the highest level of system privileges, allowing it to interact directly with core operating system functions. According to Symantec, PoisonX carried a valid Microsoft Hardware Compatibility signature, enabling Windows to load the driver as trusted.

Once active, PoisonX terminated security-related processes and removed user-mode API hooks commonly used by endpoint detection and response (EDR) solutions to monitor application behavior. By disabling those monitoring mechanisms, the attackers reduced the visibility of security software before carrying out the remaining stages of the intrusion.

With endpoint protections weakened, Hyadina expanded its operation using a collection of credential theft and reconnaissance utilities. Investigators observed fourteen open-source tools designed to recover credentials from web browsers, email clients and instant messaging applications, as well as utilities capable of extracting Wi-Fi credentials and capturing live network traffic. All but one of those tools originated from NirSoft, which publishes legitimate Windows administration utilities. The remaining tool, Mimikatz, is widely known for extracting credentials and authentication material from Windows systems and is frequently abused during post-compromise activity.

The attackers also relied on PsExec, Microsoft's remote administration utility, to move laterally across the victim's network. Combined with legitimate remote management software, credential theft utilities and the signed kernel driver, the attackers were able to strengthen their foothold across multiple systems before deploying the GodDamn ransomware payload.

Symantec also examined the origins of PoisonX. The driver was uploaded to GitHub on April 7 by a developer using the name oxfemale, who described it as a research tool. The developer regularly publishes offensive security projects, including exploit proof-of-concepts, credential stealers and software designed to disable antivirus products, while identifying themselves on LinkedIn as a Russian security researcher specializing in reverse engineering and penetration testing. Symantec, however, considers PoisonX to be malware because its primary function is to disable security protections rather than support legitimate defensive research. Researchers said it remains unclear how the driver obtained Microsoft's Hardware Compatibility signature or whether the signing process was manipulated.

Microsoft maintains a Vulnerable Driver Blocklist to prevent known malicious or exploitable drivers from loading, even when they possess valid signatures. However, Symantec noted that newly identified drivers are not added to the blocklist immediately. Updates can take days or, in some cases, weeks to reach enterprise systems, leaving a window during which attackers can continue using newly signed or newly discovered drivers before defensive protections are updated.

Commenting on the broader use of legitimate software during ransomware intrusions, Symantec's Brigid O Gorman noted that nearly any administrative or security tool can become malicious when misused. She said this makes behavioral and adaptive security controls particularly important because they focus on suspicious activity rather than relying solely on known malicious files or applications. In the case of Hyadina, that combination of trusted software, publicly available offensive tools and a signed kernel driver enabled the attackers to disable security protections, steal credentials, move laterally through the environment and ultimately deploy ransomware.

Read more »
technology
AI AI Systems Nvidia software spacex supply chain technology

Nvidia Introduces AI-Focused PC Chip as Industry Pushes Toward Local AI Processing

 Nvidia has announced a new processor designed to run artificial intelligence applications directly on personal computers, signaling the company's latest effort to expand beyond the data center market and into everyday computing devices.

The announcement was made by Nvidia Chief Executive Officer Jensen Huang during a keynote presentation in Taipei ahead of Computex, one of the world's largest technology trade shows. The new chip, called RTX Spark, was developed as part of a long-running collaboration between Nvidia and Microsoft aimed at adapting personal computers for increasingly complex AI workloads.

Unlike many current AI services that rely on cloud infrastructure to process requests, the RTX Spark platform is designed to execute AI tasks locally on laptops and desktop systems. This allows certain AI functions to operate directly on the device rather than sending data to remote servers for processing. Industry observers believe this approach could improve response times, reduce dependence on internet connectivity, and give users greater control over sensitive information.

Nvidia said the processor was developed in partnership with Taiwanese semiconductor company MediaTek. Systems powered by the chip are expected to become available later this year through several major computer manufacturers, including Dell, HP, Lenovo, ASUS, MSI, and Microsoft's Surface product line. Additional products from Acer and GIGABYTE are also expected to follow.

The launch places Nvidia in more direct competition with companies such as AMD, Intel, Apple, and Qualcomm, all of which are pursuing their own strategies for bringing artificial intelligence capabilities to personal computers. While Nvidia has established a dominant position in hardware used to train large AI models, the company is now increasingly focused on technologies that run AI applications after those models have already been developed.

A major objective behind the RTX Spark platform is support for so-called AI agents. Unlike conventional chatbots that simply answer user questions, AI agents are designed to perform sequences of tasks with limited human intervention. Potential applications include managing schedules, conducting research, organizing information, generating content, and carrying out routine administrative work.

According to Nvidia, future personal computers will need significantly more processing capability to support these systems because AI agents are expected to operate continuously in the background rather than responding only when a user initiates an action.

The company's emphasis on local AI processing reflects a broader trend emerging across the technology sector. Many firms are exploring ways to move AI workloads closer to users instead of relying entirely on cloud-based infrastructure. Supporters of this approach argue that local processing can improve performance while reducing network delays and operational costs.

The commercial success of AI-powered PCs, however, remains uncertain. Although several manufacturers have promoted AI-enabled devices as the next phase of personal computing, adoption has been uneven. Some vendors have reported positive contributions to sales, while others have indicated that demand has not reached the levels initially anticipated when the category was introduced.

Technology analysts nevertheless view the market as an area with long-term growth potential. Neil Shah, co-founder of Counterpoint Research, said the shift from application-centered computing toward AI-assisted systems could fundamentally change how users interact with their devices. He suggested that personal AI agents operating on local hardware may become increasingly common as the technology matures.

During his presentation, Huang also highlighted Nvidia's Vera central processing unit, which he previously described as providing access to a market opportunity worth approximately $200 billion. Nvidia stated that organizations including OpenAI, Anthropic, and SpaceX are among the early adopters evaluating the technology.

The Computex presentation also featured discussion about the future direction of artificial intelligence across the computing industry. Qualcomm Chief Executive Officer Cristiano Amon, speaking separately ahead of the event, argued that the industry is moving beyond AI systems that simply generate responses to prompts and toward software capable of carrying out tasks independently. He described 2026 as a potential turning point for agent-based AI, adding that existing device architectures were largely designed around actions initiated by users rather than autonomous software systems.

Huang also addressed concerns that advances in artificial intelligence could reduce employment opportunities for software developers. Rejecting that view, he argued that AI tools are increasing productivity and enabling organizations to undertake larger software projects, which in turn could create additional demand for engineering talent.

The announcements come as Nvidia continues to expand its presence across multiple segments of the AI market. After becoming one of the leading suppliers of hardware for AI model training, the company is now seeking a larger role in personal computing, inference processing, and AI applications designed to run directly on consumer devices.

The developments were unveiled in Taiwan, a location Huang described as central to the global AI supply chain. The Nvidia chief, who was born in the southern Taiwanese city of Tainan, has repeatedly emphasized the island's importance to the future development and production of advanced computing technologies.

Read more »
Subscribe to: Posts (Atom)