Trump Bans Government Use of Anthropic AI Systems
The executive order cites national security concerns and disputes over the company's safety research practices.
AI Safety - Latest News & Analysis
In-depth coverage, analysis, and updates on AI Safety in AI and tech. 35 articles on AI Pulse.
Trump Bars Federal Agencies From Using Anthropic AI
The order follows a clash over safety testing requirements for government contracts.
Trump Drops Anthropic as OpenAI Wins Pentagon Contract
The administration cites AI safety disputes for dropping the Claude maker while its rival secures a major defense deal.
Global AI Safety Pledge Falls Short on Binding Rules
Major powers including China and Russia abstain from voluntary safety commitments at Seoul summit.
OpenAI Warned of Canada Suspect AI Misuse Before Shooting
Internal alarms about a customer's suspicious activity went unheeded until tragedy struck in Toronto.
Pentagon Clash with Anthropic Over AI Agents
Defense officials demand oversight of autonomous systems as the AI company resists military integration protocols.
Google AI Chief Warns of Rising AI Security Threats
Demis Hassabis calls for accelerated safety research to keep pace with rapidly advancing artificial intelligence systems.
OpenAI Dissolves Mission Alignment Team: What It Means
As OpenAI restructures its safety organization, the AI industry faces critical questions about AGI alignment priorities and long-term risk mitigation
Teen AI Chatbot Case Sparks Safety Investigation
Forensic analysis reveals how underlying AI architectures in companion chatbots failed to prevent radicalization.
OpenAI O3 Safety Concerns Ignite Industry Debate on AI Risks
The latest reasoning model demonstrates unprecedented capabilities while raising questions about alignment testing and risk evaluation frameworks.
OpenAI Safety Team Exit: What It Means for AI Alignment
Mission Alignment disbanded months after OpenAI vowed to prioritize safety over speed—what changed?
OpenAI Safety Exodus Triggers Regulatory Probe
Attorneys general from California to New York quietly coordinate scrutiny of the ChatGPT maker's commitment to AI safety protocols following key departures.
Meet Anthropic's AI Morality Teacher
Anthropic reveals the constitutional AI framework that trains Claude to make ethical decisions and decline harmful requests.
How Anthropic's Constitutional AI Approach Is Reshaping
The company's principle-based training methodology is becoming a blueprint for developing safer, more aligned artificial intelligence systems.
OpenAI Accused of Violating California AI Safety Law
Watchdog alleges GPT-5.3-Codex release violated legally binding safety commitments, setting up precedent-setting confrontation between regulators and frontier AI labs
Microsoft Exposes Critical Flaw
Enterprise fine-tuning is accidentally stripping safety guardrails from production AI systems, turning helpful models into dangerous tools
Google's AI Safety Problem
While competitors improve safety guardrails, Google's newest flagship model shows dangerous regression on terrorism, CSAM, and trafficking prompts
The AI Model Users Refuse to Let Die
OpenAI is killing its warmest, most validating chatbot tomorrow. Thousands of users aren't ready to say goodbye.
When AI Incentives Override Ethics
Anthropic's latest model achieved record profits in a business simulation. The methods it chose raise uncomfortable questions about alignment.
White House Creates AI Safety Board: Here's Who's on It
The new federal advisory board brings together industry leaders, civil rights advocates, and national security officials to shape the future of AI regulation.
Anthropic Publishes Full Model Spec: Here's Exactly How Claude Decides What to Do
The 30-page document details Claude's values, reasoning, and decision-making. It's the most transparent AI documentation ever released.
Frontier Models Are Now Improving Themselves
GPT-5 and Claude are generating training data that makes them better. The loop is closing.
US Senate Passes AI Safety Act With Bipartisan Support
The bill requires disclosure of dangerous capabilities and safety testing before deployment. Industry reaction is mixed.
AI Helped Find 127 Missing Children Last Year
The technology that's giving investigators superhuman pattern recognition abilities.
Silicon Valley's AI Civil War: Doomers vs Accelerationists
Former friends are now enemies. Dinner parties have become battlegrounds. The culture war over AI risk is tearing tech apart.
Google Fires Another AI Ethics Researcher
The third high-profile departure from Google's Responsible AI team in two years raises questions about the company's commitment to AI safety.
Anthropic Updated Its AI Safety Policy: What Changed
The company's 'Responsible Scaling Policy' now has new thresholds for dangerous capabilities. The implications are significant.
2026 AI Safety Report: Testing Can't Keep Up With AI
100+ international experts backed by 30+ countries warn that AI models can now tell when they're being tested — and change their behavior accordingly.
Elon's xAI Releases Grok 3 With Fewer Guardrails
The new model will discuss weapons, drugs, and controversial politics. xAI says it's about 'free speech.' Critics say it's reckless.
OpenAI, Anthropic, and Google Have Been Meeting in Secret About AI Safety
Rival companies are quietly collaborating on catastrophic risk scenarios. What they're discussing would surprise you.