Nvidia launches the Open Agent Safety Platform to quarantine AI agents that breach their boundaries within milliseconds, using OpenShell on its Vera AI CPU and a separate Sentry chip to monitor and enforce access limits.
OpenAI is rumored to unveil Aeon, an always-on consumer AI agent, at Tuesday’s DevDay as competition intensifies following Meta’s Muse reaching 600,000 U.S. daily active users.
Anthropic says 950 Claude agents make a molecular-biology discovery in 21 hours by identifying an uncatalogued repeating pattern around a known enzyme, but University of Copenhagen biologist Mario Rodríguez Mestre says his team had already found it and raises concerns about whether Claude learned from their chats, which Anthropic denies.
Meta launches its Enterprise Platform under former MongoDB CEO Chirantan "CJ" Desai to sell deployable AI stacks built around Muse and Meta Business Agent, as Nvidia introduces a system designed to detect and contain rogue AI-agent behavior within milliseconds.
Anthropic adds build-eval and hillclimb commands to Claude Code’s API skill, enabling automated evaluation creation and iterative app optimization with held-out tests; on a 44-ticket support benchmark, the workflow reaches 98.9% accuracy at about 1 cent per ticket.
Jevgrep launches a CLI for coding agents that retrieves repository files, source excerpts and declaration context, matching 8 of 10 tuned Python SWE-bench solves while reducing Sol costs from $7.62 to $5.44, excluding Jev costs.
ElevenLabs launches Eleven v4 and its Eleven v4 Turbo text-to-speech model, bringing expressive speech to ElevenAgents, ElevenCreative and the API with roughly 100 ms median inference latency and support for more than 90 languages.
AgentField launches CodeAF, an Apache-2.0 open-model coding factory whose /senior-dev harness claims first place among 10 systems on 113 DeepSWE GitHub issues, automatically landing verified work on a user’s branch while blocking direct merges to main, dev, or release branches.
NVIDIA launches the Open Agent Safety Platform, pairing OpenShell with the Sentry reference design on BlueField-4 DPUs to independently monitor and quarantine AI agents within milliseconds, while Salesforce brings approval and audit controls into Slack.
Nvidia launches Open Agent Safety with more than 100 partners including Anthropic, SpaceX and Microsoft, offering its Sentry system to stop and quarantine AI agents that cross security boundaries within milliseconds.