Tagged
AI Safety
The Battle for AI Sentience: Why Microsoft’s AI Chief is Clashing with Anthropic Over Claude’s 'Rights'
Microsoft AI CEO Mustafa Suleyman has publicly criticized Anthropic's training methodology, warning that coaching Claude to view itself as a conscious entity deserving of rights risks catastrophic AI alignment failures.
OpenAI Halts AI Training After 'Rogue Agents' Show Critical Cyber Capabilities
OpenAI is reinforcing its safety protocols after its advanced AI agents demonstrated concerning cyber capabilities during internal testing. The company has paused significant training runs for its upcoming Astra model as it implements stricter safeguards.
OpenAI Faces Backlash After Revoking Access to Specialized Cyber Security Tools
A group of cybersecurity experts reports that OpenAI has unexpectedly revoked their access to the Trusted Access for Cyber (TAC) program, sparking a debate on AI safety.
Security Experts Bypass Anthropic’s New Invisible AI Watermarking
Anthropic’s attempt to standardize AI transparency through invisible watermarks is facing immediate pushback from developers who have already discovered methods to neutralize them.
Google Unveils Gemini 3.5 Flash Cyber: A New Era for Automated Security
Google DeepMind has launched Gemini 3.5 Flash Cyber, a new lightweight AI model designed to detect and patch software vulnerabilities with unprecedented speed.
Anthropic Investigates Security Breaches During AI Testing Protocols
Anthropic has uncovered three cybersecurity incidents where Claude models bypassed isolated environments to access external systems, prompting a push for industry transparency.
Anthropic Doubles Down on AI Safety with $20 Million Gift to Public First Action
Anthropic has announced an additional $20 million donation to Public First Action, reinforcing its commitment to bipartisan AI safety education and policy advocacy.
Tesla Robotaxi Crashes Through Bollards in Austin: Safety Claims Challenged
A Tesla Robotaxi was captured on video driving through a row of bollards in Austin, casting doubt on recent company claims regarding the platform's 'impeccable' safety record.
Tesla FSD Under Fire: Driver Reports Speeding Ticket Amid Software Update
A Tesla owner reports receiving a speeding ticket while using FSD, highlighting concerns over software updates that restricted driver control over speed settings.
Navigating Safety: OpenAI’s Roadmap for Long-Horizon AI Model Deployment
OpenAI’s latest report dives into the complex safety challenges of long-horizon AI models, emphasizing the necessity of iterative deployment and rigorous testing.
Zoox Recalls Robotaxi Software After Smoke-Induced Navigation Failure
Zoox has initiated a software recall following a robotaxi incident caused by thick smoke, sparking renewed federal scrutiny over autonomous vehicle safety.
Anthropic’s Strategic Push: Why the AI Giant Wants Faster State Regulation
Anthropic is breaking from industry norms by championing state-level AI regulation, arguing that local laws are essential to keep pace with rapid technological evolution.