The Archive

OpenAI highlights adoption of its products across education, agriculture, and small business in Brazil.

Anthropic and Iceland launch national AI education pilot, expanding Claude access to students.

Cognizant deploys Claude to 350,000 employees, demonstrating enterprise adoption at scale.

OpenAI launches IndQA benchmark for evaluating AI systems across 12 Indian languages and 10 knowledge domains.

OpenAI and AWS announce $38B multi-year partnership for AI infrastructure and next-generation model scaling.

Berkeley BAIR introduces divide-and-conquer RL algorithm scaling to long horizons without temporal difference learning overhead.

OpenAI expands Stargate infrastructure project to Michigan with one-gigawatt compute campus.

OpenAI releases Aardvark, an autonomous security researcher agent for finding and fixing software vulnerabilities.

OpenAI details OWL architecture enabling agentic browsing, fast startup, and rich UI in ChatGPT Atlas browser.

Google DeepMind launches AI for Math Initiative with leading research institutions to advance mathematical discovery.

Anthropic opens Tokyo office and signs AI safety cooperation memorandum with Japan AI Safety Institute.

OpenAI technical report on gpt-oss-safeguard-120b and -20b reasoning models for policy-driven safety classification.

OpenAI releases gpt-oss-safeguard, open-weight reasoning models for custom safety policy classification.

DNP deployed ChatGPT Enterprise across 10 departments, achieving 95% faster patent research and 87% automation in 3 months.

Doppel uses GPT-5 with reinforcement fine-tuning to detect deepfake and impersonation attacks, reducing analyst workload 80%.

OpenAI recapitalization strengthens governance mission with expanded resources for responsible AI development.

Microsoft and OpenAI expand long-term partnership to strengthen innovation and responsible AI progress.

Claude for Financial Services adds Excel plugin, real-time market connectors, and pre-built modeling and analysis skills.

OpenAI submits White House proposal on energy infrastructure and workforce investment for sustained U.S. AI leadership.

GPT-5 system card documents improvements in sensitive conversation handling with benchmarks for emotional reliance and jailbreak resistance.

OpenAI collaborated with 170+ mental health experts to improve ChatGPT's distress recognition and empathetic response, reducing unsafe outputs 80%.

Steuerrecht.com uses ChatGPT Business to automate tax research and accelerate legal workflow for law firms.