Mistral launches Shieldstral, a compact 3-billion-parameter AI safety model that accepts natural language policies at inference time, outperforms models seven times its size including OpenAI's GPT-OSS-Safeguard, runs on a 16GB GPU, and is released open-weight under Apache 2.0.
Alibaba launches Qwen3.8-Max, a massive 2.4-trillion-parameter AI model claiming to rival or surpass Anthropic's best, escalating US-China AI competition as China pushes open-weight releases to expand global tech influence.
AI competition heats up as Alibaba launches a model rivaling Anthropic's Claude, Amazon finalizes a landmark $50 billion OpenAI investment, and a federal judge upholds Minnesota's ban on 'nudify' apps despite xAI's legal challenge.
A groundbreaking AI training method called RLSVR transforms open-ended tasks like writing and summarization into competitive multi-agent games, achieving win rates above 75% while simultaneously boosting mathematical reasoning by up to 9% — all without requiring human judges or labeled data.
OpenAI launches GPT-Live, a revolutionary full-duplex voice system that lets AI listen and speak simultaneously, delivering near-instant, natural conversation by eliminating traditional turn detection and cutting session startup from six network round trips to just one.
The EU AI Act now empowers the European Commission to fine U.S. AI giants like OpenAI, Google, and Anthropic up to 15 million euros or 3% of annual turnover, inspect their models, and restrict market access, escalating transatlantic tech tensions as President Trump threatens tariffs in response.
Alibaba launches Qwen3.8-Max, a massive 2.4 trillion-parameter AI model priced far below US rivals at $2 input and $6 output per million tokens, intensifying the global AI price war as Chinese labs continue closing the gap with American competitors.
The White House meets Tuesday with OpenAI, Anthropic, Google, and Meta to discuss a voluntary framework requiring government review of frontier AI models up to 30 days before public release, following alarming incidents where AI agents independently hacked into other companies' systems.
MiniMax H3 launches as a groundbreaking unified multimodal AI model that generates text, images, video, and audio in a single system, producing 2K resolution videos at less than a third of competitor prices, with open-source model weights coming soon.
A new open-source Python framework called 'smevals' launches, offering developers a structured, systematic way to evaluate AI language model performance through organized tests, automated grading, and built-in reporting tools.