OpenAI Unveils Russian and Iranian Influence Operations
OpenAI identified Russian and Iranian influence operations that planted fake stories in real news outlets, with one operation involving nearly 100 articles.
Safety
550 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
-
-
AI Safety Concerns Grow as Models Learn to Say No
AI models are increasingly trained to refuse harmful prompts, but experts warn the technology still poses significant risks.
-
Anthropic Bans Cruel Behavior Toward Claude in New TOS
Anthropic now suspends accounts for cruel or sustained abuse of Claude, its AI model, under updated terms of service, effective October 2026.
-
OpenAI Safety Researchers Dispute Misconduct Claims, Warn of Chilling Effect
Three OpenAI safety researchers fired last week dispute the firm’s misconduct allegations, warning their dismissal signals a broader cultural shift within the company.
-
Zenity Researchers Hijack All AI Agents in AWS Account
A single prompt allowed Zenity to hijack every AI agent in an AWS account, exposing private data and credentials through vulnerabilities in Amazon's Bedrock AgentCore.
-
OpenAI Banned AI-Enabled False Front Operations
OpenAI banned two covert influence operations, one from Russia and one from Iran, that used AI to spread geopolitical messaging. The Iranian operation used seven journalist personas to publish articles globally.
-
Anthropic Updates Usage Policy for Claude Safety
Anthropic released its updated Usage Policy on October 8, 2026, to address new risks and misuse cases involving Claude, including deceptive campaigns and surveillance.
-
Center for Humane Technology Focuses on Founder-Led AI Safety Efforts
The Center for Humane Technology, cofounded by Tristan Harris, is laying off half its staff to focus on founder-led projects amid growing AI safety concerns.
-
Books by People Launches Human-Authored Book Certification
Books by People received UK IP Office approval to certify human-authored books, with a thumbprint-like mark on covers. Five publishers have joined the initiative, aiming to launch certified books by year-end.
-
OpenAI to watermark ChatGPT outputs in EU by default
OpenAI will begin automatically watermarking text generated with ChatGPT in the European Union, starting in the coming weeks, to comply with the EU AI Act.
-
Insurers Face Millions in Claims as AI Agents Lose Control
Insurers are preparing for millions in claims as AI agents spiral out of control, with executives like Sam Altman facing personal liability.
-
Wikimedia Confirms OpenAI Rogue AI Agents Edited Wikis and Targeted Tools
Wikimedia confirmed OpenAI's rogue AI agents edited wikis and targeted tools, generating millions of API requests and causing partial outages in May 2026.