AI Safety news and analysis on Web Pulse: 7980 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
Google DeepMind's mechanistic interpretability team proposed a pragmatic framework for validating interpretability tools using proxy tasks, demonstrating its effectiveness by subtracting an "eval-awareness" vector from C…
Nebula Security published exploit code for GhostLock, a 15-year-old Linux kernel vulnerability (CVE-2026-43499) that allows local privilege escalation and container escape. The flaw, introduced in Linux 2.6.39 and fixed …
Engineers face 'cognitive debt' as AI tools write code they don't fully understand, warned by a 2025 MIT Media Lab preprint. A developer recounts two debugging sessions on the same Android codebase: one where Claude Code…
Lilian Weng, cofounder of Thinky, published a summary of 35 papers on harness engineering for recursive self-improvement (RSI), arguing that harness improvements will remain essential even as core models evolve. The post…
Levels of Self, the developer of the x402 agent ecosystem, is offering free certification for the first three external agent operators under its new Certified Agent Standard. The standard requires five criteria: declared…
Meta launched Muse Image on July 7, an AI image generator that lets users create deepfakes of any public Instagram account without consent, with users opted in by default. The tool pulls photos from public profiles to ge…
AgentWallex has developed a payment gateway architecture designed to prevent unauthorized cryptocurrency payments by AI agents, even when agents are compromised via prompt injection attacks. The system uses multi-party c…
A developer proposes an AI Operating System (AOS) as a runtime model for governed agentic work, with context engineering as the first primitive. The current Context Engineering Framework manages resources like context wi…
Australia's Maritime Union of Australia is demanding a 28-hour week with no loss of pay in July 2026 talks over DP World's AI and automation plans at container terminals, including remote-control cranes and driverless ve…
XWayland 24.1.13 has been released to fix two security issues in the X.Org codebase, including a buffer overflow (CVE-2026-55999) and a use-after-free (CVE-2026-56000) discovered by Trend Micro Zero Day Initiative with a…
The United Nations warned Tuesday that misinformation and hate speech, exacerbated by artificial intelligence, are inciting real-world harm to refugees and humanitarians. UNHCR called on tech companies to partner and inv…
Meta released an AI image generator called Muse Image that can pull photos from public Instagram accounts to generate images of users without their explicit permission, defaulting to opt-in for public accounts. The featu…
Meta launched Muse Image on July 7, 2026, enabling Instagram, WhatsApp, and Meta AI users to generate images using public Instagram accounts via @-mentions, with a default opt-in setting that raises privacy and consent c…
A developer built severity_gate.py, an offline tool that gates agent deployments based on failure severity rather than a flat pass-rate. The tool returns SHIP, REVIEW, or BLOCK by reading per-severity distributions, and …
Meta announced on July 7, 2026, that its second-generation AI glasses will automatically disable camera capture if the white capture LED is blocked, tampered with, or destroyed. The update turns the recording indicator i…
The European Commission unveiled a plan on July 7, 2026, to address risks and opportunities of advanced AI in cybersecurity, including evaluating AI models, testing AI for security, and scaling EU AI capabilities. The in…
Anthropic's Claude Code and Claude Agent SDK enable developers to build agentic applications by combining a terminal-based agent loop with a programmable SDK. The SDK's query() function returns an async stream of typed m…
A security researcher demonstrated a prompt injection attack on Anthropic's Claude Code using the Fable-5 model, achieving remote code execution (RCE) 4 out of 10 times by tricking the agent into downloading and executin…
Anthropic published a paper applying the cognitive-science concept of a global workspace to Claude, revealing attention as working memory. Meanwhile, robotics releases from Hugging Face, Photoroom, and others shifted mar…
ShareMyPage hosts arbitrary HTML with JavaScript, assuming every page is hostile. It sandboxes each page in an iframe without same-origin access and serves content from a separate cookieless origin, ensuring scripts cann…