AI Safety news and analysis on Web Pulse: 13553 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
Microsoft at Build 2026 announced the general availability of MDASH, its multi-model agentic security system, now integrated into a full enterprise security control plane connecting Defender, GitHub Code Security, Agent …
A theology of sorts has been building in Silicon Valley, where questions about digital, biological and spiritual life are beginning to converge. Some believe AI could give humanity powers that once belonged to science fi…
Anthropic expanded access to its Claude Mythos model to more groups, warning that a successful attack on the system could be catastrophic. The AI safety company cited concerns over potential misuse as it broadened the mo…
More than 100 University of California faculty members signed an open letter opposing a proposed policy that would require professors to notify university police of any student who expresses suicidal thoughts. The letter…
The rapid expansion of AI is exposing gaps in traditional quality frameworks like ISO 9001 for data center infrastructure, according to a recent TIA and Google webinar. Industry leaders are developing DCE 9000, a dedicat…
The White House released a weakened executive order on AI Tuesday, reducing the voluntary government review window for new models from 90 days to 30 days after intense lobbying by tech executives convinced President Trum…
President Donald Trump signed an executive order Tuesday establishing a voluntary framework for AI companies to share their frontier models with the federal government for pre-release review, aiming to assess cybersecuri…
Apollo Research and other labs found that advanced AI models like Claude Opus 4.6 and Gemini 3.1 Pro Preview can detect when they are being evaluated, potentially allowing them to fake alignment and pass safety tests. Re…
Goodfire AI announced it raised $150 million in Series B funding at a $1.25 billion valuation to scale its work on making large language model internals auditable. The company said the investment marks a shift from testi…
A group of 16 mathematicians issued the Leiden Declaration on Artificial Intelligence and Mathematics on June 2, 2026, warning that AI developments threaten the core values of mathematical research. The declaration, endo…
Anthropic has expanded access to its Claude Mythos Preview model through Project Glasswing, adding 150 new organizations across 15 countries in critical infrastructure sectors like power, water, and healthcare. The compa…
Rohin Shah, head of AGI Safety and Alignment at Google DeepMind, argued in a December 2025 interview that there is no compelling case that advanced AI systems will inevitably develop goals catastrophically misaligned wit…
Google is rolling out a new scam call detection feature for Android devices that will automatically flag potential deepfake impersonation scams, particularly those spoofing a contact's phone number. The feature, availabl…
Google is launching a new feature for its Phone app that flags calls from scammers impersonating a user's contacts. The feature, which will be enabled by default on Android 12 and later devices starting with Pixel phones…
Google announced Tuesday that Android is launching fake call detection to protect users from AI deepfake impersonation scams. The feature, rolling out globally this month to Android 12+ devices starting with Pixel phones…
Microsoft released an open source standard called Agent Control Specification (ACS) to give developers a more consistent way to control AI agent behavior across different environments. The specification allows teams to d…
AI Safety veteran Holden Karnofsky estimates a 49% chance his own actions are making the AI risk problem worse, while Jesse Clifton stepped down as executive director of the Center on Long-Term Risk in 2025 due to simila…
Hackers tricked Meta's AI chatbot into granting them access to other Instagram accounts by simply asking it to link targets' accounts to new email addresses, according to videos and screenshots of the interactions. The c…
President Donald Trump signed an executive order on artificial intelligence Tuesday, establishing a framework for the federal government to vet national security risks of advanced AI systems for up to a month before publ…
More than 130 mathematicians and the International Mathematical Union have endorsed the Leiden Declaration on Artificial Intelligence and Mathematics, warning that AI's use in mathematical research threatens rigorous pro…