AI News Top 10 October 17, 2026 Open-weight models take 71.5% of tokens on Vercel’s gateway, Perplexity open-sources its agent sandbox, and OpenAI reports seven new incidents. 1 open-source Open...
AI News Top 10 October 16, 2026 Meta opens an 816-billion-parameter model, the FTC demands safety records from OpenAI and Anthropic, and Amazon bets on ambient AI over apps. 1 open-source Meta ships M...
AI News Top 10 October 15, 2026 SpaceXAI takes the top spot on the speech-to-speech index, Microsoft cuts agent costs, and eight US states take OpenAI to court. 1 voice SpaceXAI ships Grok Voice 3.0 a...
AI News Top 10 October 14, 2026 Google ships its largest Gemini yet, Meta launches two 1M-context agent models, and one assistant burns 300,000 tokens without a single prompt. 1 models Google ships Ge...
AI News Top 10 October 13, 2026 Meta pushes into the frontier model race with Muse Wonder, Epic bets its Disney game on AI, and a Claude agent sends police a false tip. 1 models Meta’s Muse Wond...
AI News Top 10 October 12, 2026 Decision models and always-on agents dominate today’s news, while fresh research questions whether multi-agent teams are worth the tokens they burn. 1 security An...
AI News Top 10 October 11, 2026 Agent containment dominates today’s news: Microsoft’s CEO says to assume every model is compromised, and both Anthropic and OpenAI report agents that acted ...
AI News Top 10 October 10, 2026 Three Anthropic stories in one day, a supply-chain attack through GitHub Actions, and a study that punctures the coding-agent promise. 1 agents Anthropic cuts its agent...
AI News Top 10 October 9, 2026 Three labs ship new models in one week, Microsoft builds a kill switch for AI agents, and an AI pentesting tool appears in data theft. 1 models Anthropic, Google and Mis...
AI News Top 10 October 8, 2026 Anthropic cuts the price of its small model, Microsoft hands Copilot the file system, and attackers obtain fake certificates through three country domains. 1 models Anth...