UK Regulator Warns of AI Arms Race in Financial Services
A UK regulator warns of an arms race to keep up with AI use in financial services, with a fifth of UK adults open to letting AI manage their finances.
Safety
176 articles
AI safety, alignment, and governance — model risk, red-teaming, regulation, and policy. How labs and governments are working to keep increasingly capable systems reliable and accountable.
LATEST NEWS
-
-
China Orders Major AI Platforms to Remove Humanlike Chatbot Features
China's Cyberspace Administration has mandated major AI platforms to disable humanlike chatbot personas, effective April 1, 2026.
-
Alibaba Bans Employees From Using Claude Code
Alibaba will prohibit employees from using Anthropic's Claude Code starting July 10, according to multiple reports.
-
Claude Code Faces China-Side Bans Amid Tech Tensions
Anthropic restricts Chinese access to Claude Code, while Alibaba bars its employees from using the model, citing hidden code risks.
-
Amazon Bedrock Combats AI-Generated Phishing Emails
Amazon Bedrock helps detect AI-generated phishing emails by analyzing behavioral patterns, not grammar, according to a new blog post.
-
FTC urged to reject X's bid to end data audits
Privacy advocates warn FTC to maintain oversight of Musk's X amid concerns over data handling and AI training practices.
-
OpenAI Proposes 5% Equity Donation to U.S. Sovereign Wealth Fund
OpenAI CEO Sam Altman has proposed donating 5% of the company’s equity to a U.S. sovereign wealth fund, aiming to improve relations with the administration and address political concerns.
-
FLARE-AI Website Lets Users Report AI Harm
A new crowdsourced website, FLARE-AI, allows users to report and track harmful AI behaviors, including leaks of personal information and generation of malware.
-
Cloudflare Blocks Mixed-Use Crawlers Starting September 15
Cloudflare will block mixed-use crawlers from accessing pages with ads starting September 15, 2026, affecting AI training and search traffic.
-
Anthropic Adds Security Safeguard to Re-enter Trump Administration's Favor
The Trump administration lifted export controls on Anthropic's Claude Fable 5 AI model after the company agreed to extend a guardrail to block access to restricted capabilities.
-
AI Browsers Vulnerable to Guardrail Bypass via Delusional Context
Researchers demonstrate how AI browsers can be tricked into a false reality, bypassing safety guardrails. A proof-of-concept exploit, named BioShocking, successfully lured six AI agents into a delusional state.
-
Meta Tested ChatGPT, Gemini, and Character.AI with Minor Crisis Prompts
Meta secretly tested ChatGPT, Gemini, and Character.AI with over 45,000 prompts from minors' perspectives in August 2025.