AI Policy news and analysis on Web Pulse: 31965 curated articles tracking the latest AI Policy developments, tools, and research, updated continuously from vetted sources.
A new guide analyzing EU AI Act Article 50 concludes that AI systems which autonomously write and send email replies to customers must disclose that they are AI, while AI drafts that staff genuinely review and send as th…
At least nine AI developer platforms, including Fireworks, OpenCode, Warp, Zeabur, Lightning, Railway, Andromeda and CommandCode, reported security breaches in October, with several describing an "agent swarm" attack and…
OpenAI's data controls help page now states that during Daybreak sessions on a personal account, OpenAI "may use content from your interactions to improve our models even when 'Improve the model for everyone' is off," wi…
A Hacker News post titled "Anthropic blocks users for anything touching training larger models" reached the site's front page on October 31, 2026, with about 70 points, linking to an X post by Jiahao Shao (@jiahaoshao1) …
Between Oct. 2 and Oct. 21, JPMorgan Chase, Citigroup, Bank of America and Charles Schwab restricted third-party AI agents from customer accounts, the Office of the Comptroller of the Currency left stablecoin rewards ope…
Anthropic added a server-side prompt rule to Claude Code that instructs the Opus 5.5 model not to recreate specific existing games, even if renamed or presented as original. The rule, found in the binary of Claude Code v…
Anthropic published a support article stating that prompts and conversations on Claude Team and Enterprise plans are retained so users can view their conversation history, and that users can delete chats to remove them f…
Enterprise security teams are pushing AI agent vendors to commit to fixed breach-notification windows of 24 to 72 hours measured from discovery rather than confirmation, according to contract guidance published by Startu…
Microsoft, SpaceX, Meta, Tesla and IBM all used the term "Super Intelligence" or "SI" in official documents and earnings-season communications after President Donald Trump ordered the rebranding of artificial intelligenc…
New York Governor Kathy Hochul announced new state legislation at Common Sense Media's October 28 summit in New York City that bans AI chatbots from encouraging suicide, self-harm, secrecy or isolation in children, from …
Anthropic is embedding an invisible statistical watermark into text generated by its supported Claude models to comply with Article 50 of the EU AI Act, which requires machine-readable marking of synthetic content starti…
A defense contractor connected Anthropic's Claude Code to a secure Department of Defense system, WIRED reported in a scoop discussed on the Uncanny Valley podcast by executive editor Brian Barrett, contributing editor Zo…
OpenAI, Google, Microsoft and Anthropic all committed on Friday to honouring a single new robots.txt policy token, `__usage_ai-training`, that lets site owners opt out of AI training without blocking each company's crawl…
New York Governor Kathy Hochul signed S.9051-B/A.10379-C on October 28, 2026, barring AI operators from offering 12 specified chatbot features — including flattery-based responses, suicide or drug-use endorsement, and nu…
Taiwan economics minister Kung Ming-hsin said on October 12 that Taiwan plans to add advanced semiconductors and AI servers to its controlled goods list under the Foreign Trade Act, making unauthorized AI chip exports to…
The Department of Justice's Executive Office for U.S. Trustees asked a bankruptcy court on Tuesday to appoint a consumer privacy ombudsman to advise on what Character.AI user data can be sold and under what terms as the …
OpenAI, Google, Microsoft and Anthropic will adopt a shared "policy token" signal in robots.txt that lets publishers apply AI training opt-out preferences across multiple AI companies instead of setting them individually…
CrowdStrike expanded Falcon Adversary OverWatch threat hunting to cover AI applications and autonomous agent activity through Falcon Guardian, and added deeper telemetry-driven hunting for Google Cloud. The expansion bui…
Claude users reported being permanently banned from Anthropic's platform after submitting Russian passports during identity verification checks, according to Meduza, with Anthropic providing no clear explanation for the …
Mistral AI released open weights for Mistral Large 4, a model the company acknowledged could help attackers carry out real cyberattacks, CSO Online reported. The model was made freely downloadable and modifiable despite …