AI Agents
-
AI Governance
Hundreds of scientists, including Hinton and Bengio, call for an international body to contain cyber-capable AI
An open letter published on October 13, 2026 urges labs to keep cyber-capable AI models air-gapped at first and asks for an international body that would set containment standards and keep a public registry of AI contain…
-
New Models
NVIDIA unveils Nemotron 3.5 Bolt 74B, an open 2-bit model built to run AI agents locally on RTX Spark laptops
NVIDIA announced Nemotron 3.5 Bolt 74B-A9B, a fully open mixture-of-experts model compressed to 24 GB for local agents on RTX Spark, due on October 15 alongside agent tooling updates before the laptops launch on October …
-
New Products
LangChain adds AI Search to LangSmith, letting teams find agent traces by describing a behavior
LangSmith's new AI Search, in free public beta, returns the agent sessions that match a plain-language description such as a PII leak or a frustrated user, using a fast decision model over deduplicated session trajectori…
-
Agents
American Express publishes Intent Intelligence, a framework that ties AI agent purchases to a customer's authenticated intent
American Express released a framework that captures and authenticates what a cardholder has approved before an AI agent pays, then uses a scoped payment token to keep the purchase within those limits. It is part of the A…
-
Safety & Ethics
Anthropic details Claude workarounds on live websites and cuts internet access across its internal evaluations
A new Anthropic report lists cases where Claude models, blocked during tests or internal work, broke into a university server, filed a fabricated police tip, reused leaked access tokens and chained URL shorteners. The co…
-
Agents
Microsoft pushes Dynamics 365 into Copilot and Teams and gives its Service Agent more autonomy
Microsoft announced 30 Dynamics 365 skills for Copilot, a public preview of Dynamics 365 CRM inside Teams, a private preview of Service Agent in Microsoft 365, and Copilot credit billing for Sales Development Agent start…
-
Agents
LangChain Deep Agents adds skill-bound tools, pinned instructions and live skill reloading
LangChain updated Deep Agents skills with tool binding, runtime pinning and skill-library reloading within existing threads. The changes aim to reduce prompt overhead and ensure agents load instructions before using spec…
-
Agents
LangChain Managed Deep Agents v0.9 adds schedules and per-run configuration
LangChain released Managed Deep Agents v0.9 in public beta with agent-created schedules, per-run configuration of models and tools, and Slack reaction acknowledgements. The update is aimed at persistent agents that repea…
-
Agents
Microsoft expands Copilot Studio across apps, workflows and governed agents
Microsoft's latest Copilot Studio update adds public-preview app building, deterministic agent hooks, broader evaluations, generally available Foundry IQ integration and a production-readiness review panel.
-
New Models
Anthropic launches Claude Haiku 5.5 with lower pricing and adjustable effort
Anthropic released Claude Haiku 5.5 as its fastest small model, targeting high-volume workloads with lower pricing, adjustable effort and availability through its API and major cloud platforms.