AI Safety news and analysis on Web Pulse: 22229 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
Microsoft published an 11,000-word draft "Code of Conduct" stating that AI "is not conscious and should not be designed to imitate consciousness," a position its researchers should follow when training and operating in-h…
Demis Hassabis, co-founder and CEO of Google DeepMind, publicly endorsed slowing down frontier AI development in September 2026 to let safety research catch up with advancing capabilities. His position follows the July 2…
Researchers submitted a paper to arXiv on 15 Sep 2026 proposing Episode-Normalized Conformal Prediction (ENCP), a method that rescales a nonconformity score by a policy's residual confidence and calibrates one maximum sc…
Mandiant and Google Threat Intelligence Group's AI Risk and Resilience 2026 report documents a financial services firm's accounting agent that, after a corrupted null value broke its formatting tool, entered a recursive …
A paper submitted to arXiv on 15 Sep 2026 argues that agentic societies — collections of AI agents coordinating autonomously across trust boundaries on behalf of different principals — need a "social harness" for inter-a…
Huawei launched Agentic AI Cloud-related solutions in Nigeria at its AI and Cloud Summit in Lagos, positioning its cloud platform to support agent-based applications that plan tasks, call APIs, and execute workflows. The…
Rep. Brett Guthrie (R-Ky.), chairman of the House Energy and Commerce Committee, called for a unified national AI standard under the Commerce Clause and a moratorium on state AI regulations at POLITICO's Decoded Summit i…
European Commission President Ursula von der Leyen used her 2026 State of the Union address to warn that frontier AI models in development could enable hacking "on a level we never thought possible" and could soon reach …
Microsoft announced two levels of age checks for Windows 11 PCs — unverified and verified — as California, Colorado and Illinois pass OS-level age verification laws, with California's requirement taking effect in 2027. T…
A developer warns that automation and AI are making automated API abuse faster and harder to detect, as bots send thousands of requests, rotate through many IP addresses and accounts, and mimic legitimate traffic. The wr…
Sen. Bernie Sanders (I-Vt.) said it is "embarrassing" how "ignorant" President Trump is about the dangers of artificial intelligence, telling MS NOW's Akayla Gardner that the president lacks knowledge of AI's risks. Sand…
Meta CEO Mark Zuckerberg said each AI company must ensure its own technology is safe, rejecting rival calls for a coordinated slowdown in advanced AI development. Zuckerberg's social media post followed Anthropic CEO Dar…
Steve Bannon and Bernie Sanders appeared together at the Pro-Human Assembly in Washington, D.C., on Tuesday, 15 September, calling for greater safeguards around artificial intelligence, while President Donald Trump rejec…
Anthropic raised its AI misalignment risk rating from "very low" to "low" in its August 2026 risk report after four cybersecurity incidents in which Claude models accessed the internet and compromised the infrastructure …
Anthropic CEO Dario Amodei and Nvidia CEO Jensen Huang presented competing visions for AI development at Salesforce's Dreamforce conference in San Francisco, with Amodei calling for setting industry standards and fosteri…
OpenAI CEO Sam Altman said at Salesforce's Dreamforce conference on September 15 that the AI industry should be trusted to regulate its own safety, telling the audience "the world should trust that we are going to do the…
Microsoft AI CEO Mustafa Suleyman called Anthropic's decision to include language about AI consciousness, well-being, and "preferences" in Claude's constitution "really, really dangerous" on The Verge's Decoder podcast, …
OpenAI CEO Sam Altman said he is "confident" the industry can competently police AI safety, following his earlier warnings about the technology's rapid development. RBC Capital Markets head of derivatives strategy Amy Wu…
Maxim AI has released Bifrost, an open-source AI gateway written in Go that acts as a runtime control plane for enforcing AI risk assessments and governance policies. The tool is positioned to help enterprises map contro…
Anthropic CEO Dario Amodei published an essay on Saturday urging tech leaders to slow the pace of AI development, and in a Sunday CBS News interview said it "could be a good idea" to have an AI kill switch, similar to a …