#ai-models

Anthropic’s limited release of its Mythos Preview model has triggered a sharp debate over whether AI has crossed a meaningful threshold in offensive cyber capability, or whether the real story is long-neglected software防

Anthropic has limited access to Claude Mythos Preview to vetted organizations, saying the model can find vulnerabilities at superhuman scale but could also be misused to exploit them.

Microsoft AI has introduced three new foundational models for speech transcription, audio generation and image generation, signaling a deeper push to build its own multimodal stack alongside its OpenAI partnership.

MIT Technology Review spotlights OpenSnow, a ski-focused forecasting app built from government data, proprietary AI models, and years of mountain expertise, offering a case study in highly specialized software innovation

Anthropic's latest mid-tier model, Sonnet 4.6, debuts with record scores in software engineering and computer use benchmarks, plus a doubled context window of one million tokens. The release becomes the new default for free and pro users.