Yasna brings together recent headlines from selected sources and makes them easier to sort with tags, filters, and search.
Search across headline titles and summaries.
Weekly headline count for the current query.
Microsoft researchers have detailed an exploit chain, named AutoJack, that turns an AI browsing agent into a delivery vehicle for remote code execution
AutoJack is a novel exploit chain showing how a single malicious webpage can turn an AI browsing agent into a remote code execution vector on the host machine. By abusing trust in localhost, missing authentication, and unsafe parameter handling, attackers can trigger arbitrary process execution through AutoGen Studio’s MCP WebSocket. The research highlights a broader pattern - when agents can browse untrusted content and access local services, traditional boundaries like localhost are no longer secure. The post AutoJack: How a single page can RCE the host running your AI agent appeared first on Microsoft Security Blog.
Cybersecurity researchers have discovered a remote denial-of-service exploit that affects major web servers, including NGINX, Apache HTTPD, Microsoft IIS, Envoy, and Cloudflare Pingora
Mythos Moves the Needle on AI Innovation, DefenseAnthropic’s “Mythos moment” is accelerating vulnerability discovery, but speed without validation is a growing risk. As exploit windows shrink and remediation lags, more findings only mean more noise. The real advantage lies in validating what actually matters—and fixing it first.
Outsiders Could Exploit Misconfig to Stream Commands, CredentialsA misconfiguration in Microsoft's Azure SRE Agent may have allowed any Azure account holder from any company to tap into another organization's agent conversations in real time, watching commands, outputs and credentials, leaving no trace.
Commands Push Lasting Preferences Into AI AssistantsMicrosoft researchers found companies embedding hidden commands in "summarize with AI" buttons to plant lasting brand preferences in assistants' memory. The tactic, dubbed AI recommendation poisoning, exploits persistent memory features to bias future responses.
This week saw a lot of new cyber trouble. Hackers hit Fortinet and Chrome with new 0-day bugs. They also broke into supply chains and SaaS tools. Many hid inside trusted apps, browser alerts, and software updates
The recent mass-theft of authentication tokens from Salesloft, whose AI chatbot is used by a broad swath of corporate America to convert customer interaction into Salesforce leads, has left many companies racing to invalidate the stolen credentials before hackers can exploit them. Now Google warns the breach goes far beyond access to Salesforce data, noting the hackers responsible also stole valid authentication tokens for hundreds of online services that customers can integrate with Salesloft, including Slack, Google Workspace, Amazon S3, Microsoft Azure, and OpenAI.
A zero-click exploit called EchoLeak reveals how AI assistants like Microsoft 365 Copilot can be manipulated to leak sensitive data without user interaction. This entry breaks down how the attack works, why it matters, and what defenses are available to proactively mitigate this emerging AI-native threat.
Attackers aren’t waiting for patches anymore — they are breaking in before defenses are ready. Trusted security tools are being hijacked to deliver malware. Even after a breach is detected and patched, some attackers stay hidden
They're good at zero-day exploits, too Silk Typhoon, the Chinese government crew believed to be behind the December US Treasury intrusions, has been abusing stolen API keys and cloud credentials in ongoing attacks targeting IT companies and state and local government agencies since late 2024, according to Microsoft Threat Intelligence.…
This week, a 23-year-old Serbian activist found themselves at the crossroads of digital danger when a sneaky zero-day exploit turned their Android device into a target. Meanwhile, Microsoft pulled back the curtain on a scheme where cybercriminals used AI tools for harmful pranks, and a massive trove of live secrets was discovered, reminding us that even the tools we rely on can hide risky
Threat hunters have shed light on a new campaign targeting the foreign ministry of an unnamed South American nation with bespoke malware capable of granting remote access to infected hosts
Attack Method Exploits RAG-based Tech to Manipulate AI System's OutputResearchers found an easy way to manipulate the responses of an artificial intelligence system that makes up the backend of tools such as Microsoft 365 Copilot, potentially compromising confidential information and exacerbating misinformation. Researchers called the attack "ConfusedPilot."
A server-side forgery (SSRF) bug in Microsoft's tool for creating custom AI chatbots potentially exposed info across multiple tenants within cloud environments.
Microsoft's announcement of the new AI-powered Windows 11 Recall feature has sparked a lot of concern, with many thinking that it has created massive privacy risks and a new attack vector that threat actors can exploit to steal data. [...]
Project behind the Rust programming language asserted that any calls to a specific API would be made safe, even with unsafe inputs, but researchers found ways to circumvent the protections.
A now-patched security flaw in the Microsoft Edge web browser could have been abused to install arbitrary extensions on users' systems and carry out malicious actions. "This flaw could have allowed an attacker to exploit a private API, initially intended for marketing purposes, to covertly install additional browser extensions with broad permissions without the user's knowledge," Guardio