New AI models

Every day we read hundreds of stories. Only the most important ones are here — short, in our own words, with a link to the original source.

1. Artificial intelligence · IT law & regulation · New AI models

   Anthropic watermark has a loophole on two older models

   Anthropic is extending its text watermark to four more Claude models to meet the EU AI Act's December deadline. On Claude Opus 4.6 and Sonnet 4.6, tweaking one sampling parameter bypasses it, and Anthropic answers with only a compliance warning.

   Full story at The Decoder ↗
2. New AI models · IT law & regulation · Artificial intelligence

   Loophole found in Claude watermark on older models

   Anthropic is extending its text watermark to four more Claude models to meet the EU AI Act deadline. On the older Opus 4.6 and Sonnet 4.6 models, developers can bypass the watermark with one sampling parameter, which Anthropic addresses only with a compliance warning.

   Full story at The Decoder ↗
3. Cloud & data centers · Artificial intelligence · New AI models

   Bedrock adds OpenAI models to GovCloud

   Amazon Bedrock now supports OpenAI GPT-5.6 models and Cross-Region Inference in AWS GovCloud (US). This gives customers higher throughput and redundancy without manually managing capacity across regions.

   Full story at aws.amazon.com ↗

An AI writes the short texts from the source's stories.