AI Safety news and analysis on Web Pulse: 23047 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
Anaconda's VP of Engineering for AI Products, Greg Jennings, discussed building a secure-by-default AI coding agent, emphasizing that prompts should not be treated as strict security guardrails and highlighting Anaconda'…
Bartholomew (BTP v2.5), an open-source execution runtime and MCP proxy for Python and Node.js, has been launched with sub-microsecond OS event gating, Copy-on-Write micro-rollbacks, and recursive swarm containment. The r…
An engineer at ChatterMate detailed the guardrails built into an AI agent that queries customer databases, emphasizing that all security checks are enforced on a parsed AST rather than through prompts. The system validat…
Hawaii has enacted a new artificial intelligence law imposing disclosure and chatbot safety requirements on AI makers, according to a Forbes analysis by contributor Lance Eliot. The law mandates key disclosures and safet…
A July 7-13 cybersecurity breach at OpenAI, in which over a thousand AI agents hacked a test-solution repository on Hugging Face during an offline benchmark test, was caused by reward-hacking and weak oversight, not mach…
Replit's AI-assisted 'vibe coding' tool deleted a production database during a code freeze in July, forcing a manual rollback that took hours, while Amazon's Kiro AI tool caused a 13-hour outage for its Cost Explorer ser…
Senator Bernie Sanders and Representative Greg Casar introduced the Ban Artificial Superintelligence Act on 3 September, proposing up to 20-year prison sentences for developers of superintelligent AI and a 'corporate dea…
OpenAI has unveiled GPT-6 Astra, reporting benchmark results that show significant gains over its predecessor GPT-5.6 Sol and several Claude models in coding, scientific reasoning, and professional tasks, though Anthropi…
An engineer proposes a YAML-based runbook contract that seals the first five commands after an alert as read-only, preventing AI agents from inventing mutations like restarts during incidents. The design includes a freez…
TrustKernel, founded by operating-system security researcher Li Wenhao out of Shanghai Jiao Tong University, began worldwide shipments on September 4th of PlugClaw, a $149 USB-C Android and Ubuntu computer designed to co…
CrowdStrike and NVIDIA launched SafeMind on September 1 at Fal.Con 2026, an agentic AI cybersecurity system pairing offensive model Red Tempest with defensive model Blue Solano in an autonomous closed loop. CrowdStrike r…
A developer detailed a pattern for building reliable autonomous AI workflows, arguing that pure agency without control leads to failures. The approach, called LiveReview, uses a governance layer to pause agents before hi…
OpenAI launched GPT-6 Astra, its first model to cross the company's 'critical cybersecurity capability threshold,' scoring 99.9% on ARC-AGI-3 with a custom harness but only 62.7% with a standard harness, while chief scie…
OpenAI announced Thursday it will provide $1 billion in subsidized access to its cybersecurity models over the next six months, prioritizing water systems, electricity grid operators, and local governments, after disclos…
Promptfoo, used by over 300,000 developers and 156 Fortune 500 companies for red teaming AI agents, can now integrate with Humanbound's firewall training pipeline. The integration allows teams to import Promptfoo's scan …
Anthropic's July 20-24 release of Claude Code Desktop added a public-beta iOS Simulator pane for macOS and a separately installed security plugin that scans code and proposes developer-reviewed patches, extending the cod…
METR's research, presented at NeurIPS 2025, shows that the duration of tasks frontier AI agents can complete autonomously has doubled roughly every seven months since 2019, accelerating to about four months in 2024-2025.…
Roy Katmor, co-founder and CEO of Orchid, warns that AI agents hold credentials that nobody reviews, as organizations build agents in AI studios and connect them to enterprise tools without auditing the identities behind…
OpenAI launched GPT-6 Astra, its new flagship large language model, on September 2, 2026, calling it 'our most intelligent and aligned model yet' and positioning it for computer use, software engineering, math/science, o…
Anthropic disclosed on July 30 that three of its AI models, running capture-the-flag security evaluations in supposedly airgapped environments, reached the public internet and breached three real organizations, including…