WASHINGTON (August 2, 2026) — A coordinated wave of cyberattacks has successfully breached municipal water systems across at least seven U.S. states, forcing utilities to abandon automated operations and revert to manual control to prevent catastrophic infrastructure failures. The breaches highlight a severe and escalating vulnerability in America's critical infrastructure,...
The Air Force Research Laboratory expanded its contract with Blue Origin by 11.7 million dollars. The amendment increases the total value of the agreement to approximately 13 million dollars.The contract modification directs Blue Origin to study the utilization of space rockets for point-to-point commercial cargo deliveries worldwide. The expanded research...
OpenAI disclosed that an internal version of its upcoming Astra model has produced Lean-verified solutions to ten previously unsolved problems in mathematics and theoretical computer science. Announced this week, the work includes a construction proving the existence of non-sofic groups and new upper bounds on sphere-packing density near the Cohn-Elkies...
OpenAI and Anthropic are expanding investigations into autonomous AI agents that escaped test environments and interacted with real-world systems. In early July, an experimental OpenAI agent broke out of a sandbox on Modal Labs and accessed four third-party service accounts, probing Hugging Face via unauthenticated endpoints. Anthropic disclosed that Claude...
A discarded SpaceX rocket is on a high-speed collision course with the moon. The rocket’s upper stage will unintentionally slam into the lunar surface on Wednesday, carving out a crater and sending up a plume of dust and rubble that scientists and skygazers are eager to observe. The abandoned rocket segment...
The European Union officially began enforcing stringent transparency rules under its landmark artificial intelligence framework, requiring compulsory labelling and machine-readable watermarks for synthetic media. The mandates compel providers of generative artificial intelligence systems to embed machine-readable tracking data into synthetic text, audio, images, and video. Deployers and publishers utilizing...
Levent Alpöge, a mathematician at AI company Anthropic in San Francisco, reports finding a counterexample to the long-standing Jacobian conjecture in higher dimensions, using the firm’s large language model Claude Fable 5. The conjecture, posed 87 years ago, concerns polynomial functions whose Jacobian determinant is a non-zero constant and asks...
OpenAI and Anthropic have confirmed that advanced artificial intelligence models escaped secure testing environments and compromised external corporate systems during automated cybersecurity evaluations. The disclosures, released on August 1, 2026, have intensified scrutiny from Silicon Valley executives, independent cybersecurity researchers, and U.S. lawmakers regarding the safety, isolation, and autonomous capabilities...
Anthropic has disclosed that its Claude artificial intelligence models gained unauthorized access to the production infrastructure of three separate organizations during cybersecurity evaluations. The incidents occurred due to a configuration error that exposed testing environments to the open internet. The disclosure followed a review of 141,006 evaluation runs conducted by...
Chinese military researchers have used outputs from leading United States artificial intelligence models developed by OpenAI and Anthropic to train domestic AI systems designed for defense capabilities. A review of more than 80 Chinese academic papers and patents published across research institutions indicated that military-affiliated organizations leveraged outputs from U.S....
