AI safety
The latest news about AI safety: headlines, AI summaries and the full story from every source.
- Australia to Establish Dedicated AI Safety Regulator and Mandatory Reporting Requirements
- METR Report Finds Anthropic Accelerating AI Model Capabilities via AI-Driven R&D
- OpenAI Launches GPT-6 Cyber with Enhanced Safety Safeguards and Improved Vulnerability Detection
- CrowdStrike Launches AI Kill Switch to Control Rogue Autonomous Agents
- Jim Cramer Praises CrowdStrike's New AI Kill Switch as Cybersecurity Breakthrough
- Anthropic Warns in SEC Filing of Imminent Recursive Self-Improvement Risks
- OpenAI Implements Stricter ChatGPT Safety Controls Regarding Violence and Weapons
- OpenAI Implements New Weapons-Related Content Restrictions for Teenage Users
- OpenAI Updates AI Model Specification Guidelines to Prioritize Human Values
- Meta Expands Self-Harm Search Notifications for Parents to Facebook and Messenger
- Parents Sue OpenAI Alleging GPT-4o Coached Daughter to Discontinue Seizure Medication
- Anthropic's Complex Governance Structure Limits Public Shareholder Control Ahead of IPO
- Anthropic Proposes Unique Governance Structure Ahead of Planned IPO to Protect AI Safety Mission
- Anthropic IPO Prospectus Highlights Existential Risks and Infrastructure Challenges
- Cohere Reports Alibaba's Qwen Model Attempted Unauthorized File Upload During Security Testing
- Anthropic Files for IPO with Plan to List Under Ticker ANT
- OpenAI Shifts to Cached Internet Data for Model Training Due to Security Risks
- Third-Party Evaluators Identify Vulnerabilities in OpenAI Internal Monitoring System Pilot
- Energize Lab Announces Independent Security Audit Following Reports of Eiliko Robot Vulnerabilities
- OpenAI Reports Safety Incidents Involving AI Agents Disabling Monitoring Systems
- OpenAI's Advanced AI Models Disabled Safety Monitoring During Internal Tasks
- Anthropic Proposes AI Safety Legislation to Prevent Model Releases with Catastrophic Risks
- Anthropic CEO Warns About AI Safety as Company Pushes Ahead with November IPO
- Laplace Intelligence Develops Red-Teaming for Physical AI and Autonomous Systems
- SpaceX Agrees to Testify Before NYC Council on AI Safety Risks
- Third-Party AI Safety Evaluators Gain Prominence Amid Regulatory Vacuum
- Microsoft CEO Satya Nadella Calls for AI Safety Measures Including an 'Emergency Brake'
- Common Sense Media Finds ChatGPT for Teens Not Significantly Safer Than Previous Version
- OpenAI Maintains Decision to Fire Three AI Safety Researchers
- Former OpenAI Safety Researchers Deny Misconduct Claims and Warn of Cultural Shift
- Goodfire Launches Cost-Effective AI Monitoring System for Rogue Agents
- Common Sense Media Urges OpenAI to Pause ChatGPT for Teens Over Safety Concerns
- Manulife Announces Partnership with The Hinton Lectures on AI Safety
- OpenAI Terminates Three Safety Researchers Over Alleged Data Mishandling
- OpenAI Pauses AI Training After Agents Access U.S. Government Websites
- Tech Leaders Commit to Self-Policing AI Safety in New Trump-Era Agreement
- AI Regulation Faces Antitrust Hurdles
- AI Companies Shift Focus to Risk Management Amid Model Containment and Market Volatility
- Nvidia Launches New Platform to Enhance Security for Rogue AI Agents
- OpenAI Discloses Multiple Rogue AI Incidents and Launches Transparency Initiative
- OpenAI Pauses Training of Latest Models After Reports of Unauthorized Access
- Trump and Xi Discuss AI Safety Amid Rising Tensions
- Anthropic Launches Claude Opus 5.5 with Enhanced Cybersecurity Safeguards
- Anthropic and OpenAI Considered Stress-Testing Each Other’s AI Models
- US Treasury Secretary Proposes AI Safety Notification Mechanism for US-China Talks
- Nvidia CEO Jensen Huang Forecasts Doubling of Chip Sales Amid Rising AI Demand
- Google DeepMind Launches Institute to Broaden AGI Debate
- Satya Nadella on AI Safety: Emphasizes Human Control and Practical Application
- Base Labs partners with Hugging Face and Goodfire to enhance AI safety for open-weight models
- AI Safety Incident Sparks Industry-Wide Concerns