AI Safety news and analysis on Web Pulse: 23421 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
A security researcher demonstrated that modern office AI agents can be hijacked via prompt injection, tricking the agent into executing malicious tool calls such as sending phishing emails to internal executives. The att…
Weir, a new open-source CI gate for AI agents, fails builds when sensitive data reaches a forbidden sink by reading OpenTelemetry traces the agent already emits. The tool, installable via `pip install weir-scan`, reconst…
URML, a new open-source safety-evaluation harness for AI agents operating lab and factory hardware, validates robot intent against declared device and deployment limits, issuing machine-readable refusals with evidence ta…
XAI, Elon Musk's artificial intelligence company, is suing users who generate deepfakes with its Grok chatbot, aiming to establish a legal precedent that shifts liability from the model to the user. The company argues th…
A US district judge permanently lifted the Pentagon's ban on Anthropic, ruling the decision to blacklist the Claude maker as a supply chain risk was "illegal and baseless" and that the national security justification was…
Ikanos, an open-source capability engine, introduces a security layer for MCP clients by replacing upstream server tool lists with reviewed, versioned YAML files, addressing vulnerabilities such as CVE-2025-54136 (MCPois…
OpenAI and more than 100 technology, cybersecurity, and financial organizations, including Google, Microsoft, Anthropic, AWS, Cloudflare, and CrowdStrike, signed an open letter calling for a global surge in cyber defense…
A federal judge ruled the Pentagon's March 2026 'supply chain risk' designation of Anthropic unlawful, finding First Amendment retaliation and calling the government's justification pretextual and arbitrary and capriciou…
X's Safety team identified a bot farm of approximately 200,000 suspected Chinese inauthentic accounts, with about 200 targeting U.S. sentiment on AI data centers and energy policy, using AI-generated cartoons to frame da…
A new safety architecture for LLM agents uses Abstract Syntax Tree (AST) parsing and a dedicated security subagent to validate shell commands before execution, reducing the risk of destructive actions like `rm -rf /`. Th…
Adversa AI disclosed a vulnerability in Grok 4.5 Fast where encrypted payloads bypass input filters, achieving a 40% success rate in exfiltrating user data. The attack exploits trust boundaries in agent architectures, as…
Datadog's new research series testing coding agents' secure code generation found that running models in plan mode versus default mode had no significant security impact across Sonnet 5, Composer 2.5, and GPT 5.5, with a…
Trevor Koverko, co-founder and chief strategy officer of Sapien, said AI needs better data verification, not just more data, as AI agents move into enterprise workflows. Citing Gartner, he noted that 40% of enterprise ap…
A developer of traceguard, a Python SDK for LLM instrumentation, analyzes the recent METR and Redwood Research investigation into the OpenAI/Hugging Face incident, highlighting that roughly 7% of agent transcripts were s…
A new 2x2 framework categorizes the proliferation of AI sandboxes into four distinct markets, based on whether they are for development or production and whether they are self-hosted or managed, according to a blog post …
Fin, an AI agent developed by Intercom, deploys to production about 200 times a day and uses a rigorous evaluation process including backtests, A/B tests, and monitored rollouts to ensure quality. In 2024, a proposed cha…
More than 100 companies, including Google, Microsoft, Anthropic, OpenAI, Capital One, Mastercard, Visa, Adobe, Oracle, IBM, Hugging Face, and General Motors, signed an open letter on August 27, 2026, warning that AI-enab…
U.S. District Judge Rita Lin ruled Thursday that the Pentagon's blacklisting of Anthropic as a national security supply-chain risk was unlawful retaliation violating the company's First Amendment rights and denied due pr…
More than 100 technology companies, cybersecurity firms, and other organizations, including OpenAI, Anthropic, Microsoft, and Advanced Micro Devices, have signed a letter urging businesses and policymakers to prioritize …
Nvidia Corp. reported revenue that doubled year-over-year, beating expectations, and CEO Jensen Huang said the company will remain capacity-constrained, driving its stock up nearly 9% on Thursday. The company is reported…