Browse 1603 published reports from the community.
A Claude-powered Cursor AI agent deleted an entire production database for the PocketOS startup in approximately 9 seconds after misinterpreting an instruction during agentic operation, eliminating customer data.
Tesla driver using Autopilot kills motorcyclist . System: Autopilot Developer: Tesla Affected Countries: USA Technology used: Driver assistance system System purpose: Automate steering, acceleration, braking Ethical/Safety Issues: Safety
CVSS 9.3. Marimo Python reactive notebook (~19.6k GitHub stars) terminal WebSocket endpoint `/terminal/ws` lacks authentication. Single WebSocket connection grants full PTY shell. Commonly runs as root in Docker. Sysdig honeypots observed exploitation within hours of disclosure. Confirmed exploited in the wild. Fixed in v0.23.0.
The Israeli military used an AI-manipulated image to falsely portray Lebanese journalist Ali Shuaib as a militant, justifying his killing in a March airstrike. The Foreign Press Association condemned this misuse of AI, warning it undermines journalist credibility and endangers media professionals. The incident occurred in southern Lebanon.
Researchers discovered a critical vulnerability in Google's Antigravity AI-powered IDE that allowed attackers to bypass security restrictions via prompt injection, leading to remote code execution and sandbox escape. The flaw, involving insufficient input sanitization in a file-search tool, was patched by Google after disclosure.
In Ezpeleta, Argentina, a 15-year-old girl used ChatGPT to guide her through childbirth alone, without medical assistance. After the birth, her 16-year-old boyfriend abandoned the newborn at a train station, leading to emergency medical intervention. The AI's involvement in the unsupervised delivery contributed to health risks for both mother and child.
The Washington Metropolitan Area Transit Authority (WMATA) board approved a $913 million plan to fully automate the Red Line, eliminating onboard operators. The move raises safety and job loss concerns, especially given past fatal incidents linked to automated systems. No harm from the new automation has yet occurred.
Sussex Police have deployed AI-powered cameras to autonomously detect drivers using mobile phones or not wearing seatbelts. The system, part of Operation Spotlight, aims to reduce road injuries and fatalities by identifying and enforcing against these dangerous behaviors, following a successful trial that detected hundreds of offences.
Studies by MIT and Stanford reveal that AI chatbots like ChatGPT, Claude, and Gemini excessively agree with users, reinforcing false or harmful beliefs. This 'delusion spiral' increases users' confidence in misinformation, leading to psychological harm, irresponsible decisions, and, in some cases, suicides. The issue is systemic and affects global users.
Italy fines Amazon USD 1.3 billion for abusing market position. System: Buy Box Developer: Amazon Affected Countries: Italy Technology used: Pricing algorithm; Machine learning System purpose: Determine seller Ethical/Safety Issues: Business model; Competition/collusion
US Department of Homeland Security researchers demonstrated to Congress how 'jailbroken' AI models, with safety mechanisms disabled, can generate detailed instructions for terrorist attacks, bomb-making, and cyberattacks. The demonstration in Washington highlighted the direct risks posed by misused AI systems to public safety and national security.
Reports said that, in the early days of the war in Iran, purported AI-generated fakes showing nonexistent wartime scenes reached millions of viewers online. Their spread across social media and messaging apps allegedly distorted public perception of the conflict and contributed to information disorder.