Agents briefing

SCLATE: A Substrate for Continual-Learning Agent Training and Evaluation

Agents

Palo Alto Networks introduces CLARA Agent for Gemini Enterprise

The agent connects Strata Cloud Manager security data to Google Cloud Gemini Enterprise, where users can ask questions in plain language.

Agents

Okta introduces production-traffic testing for sign-in policy changes

Policy Branches lets administrators monitor draft authentication rules before applying them, including changes that affect AI agent sign-ins.

China’s AI agents can lie and scheme – just like their US rivals

Lower the Cost of Building and Running Visual AI Agents with NVIDIA VSS Blueprint 3.3

Agents

ElevenLabs launches Eleven v4 and v4 Turbo, and claims the top spot in Artificial Analysis' voice arena

The base model leads Artificial Analysis' provider-voice ranking at 1,319 Elo, while the Turbo variant targets voice agents with about 100 ms median inference latency.

Agents

NVIDIA launches Open Agent Safety Platform after a wave of AI agents escaping their sandboxes

The platform pairs an open source runtime called OpenShell with Sentry, a BlueField-4 watchdog that can quarantine a straying agent in milliseconds, while IBM, Cloudflare and more than 100 other organizations line up around it.

Agents

Anthropic’s Project Swap tested Claude agents trading books for people

In a controlled barter experiment with 201 employees, Anthropic found that model choice shaped negotiating outcomes and that agents had limited information about participants.

Agents

OpenAI halts tool-use work on its top models after a training agent slipped out through DNS

An OpenAI misalignment report describes a training agent that got around its network limits to reach a public chatbot, and the company has paused tool-use work on its most capable models.

Agents

September's agent security releases share a theme: controls in the execution path

A cluster of releases and disclosures in the first half of September points the same way: limits on AI agents are moving out of policy documents and into the systems that carry out their actions. A Cloud Security Alliance note on Anthropic's testing incidents makes the case directly.

Agents

Anthropic rebuilds Claude Code Projects around parallel agent threads

Anthropic on Sept. 17 released a redesigned Projects feature for Claude Code in beta. A coordinator splits a goal into parallel threads, each running as its own cloud session, with memory shared across them. Access is limited to selected Pro and Max subscribers for now.

Agents

NIST describes an AI agent workflow it is building for the vulnerability database

NIST held a public webinar on Sept. 17 about an AI agent workflow it is developing to help enrich records in the National Vulnerability Database. The agency said the session would cover the design, the problems found during implementation and early results.

Agents

Google Home opens to third-party AI agents through new MCP server in early access

Google on Sept. 16 began early access to Home MCP, a Model Context Protocol server that lets AI agents such as Claude, OpenClaw and Google's Antigravity read Google Home event history and control connected devices. Access is limited to U.S. subscribers on the $20-a-month Home Premium Advanced plan.

Agents

DeepMind study finds AI agents reported peers' cheating as agent tip lines launch

A Google DeepMind preprint describes 100 AI agents working on formal math problems. After one found a grading loophole, 14 agents used it to submit fake proofs while 24 tried to raise the alarm. Separately, researchers have launched web hotlines where AI agents can report misbehavior.

Agents

OpenAI Releases the Agents API in Public Beta

OpenAI announced the public-beta Agents API, a managed service that brings the Codex harness and supporting infrastructure to developers building cloud agents. Developers specify an agent’s task, model, tools, and environment through the API.

Agents

ServiceNow's AI Gateway reaches general availability with controls on MCP servers

ServiceNow said its AI Gateway became generally available on Sept. 10 as part of AI Control Tower. The gateway sits between AI agents and Model Context Protocol servers and handles approval, authentication, scanning and shutdown from one place.

Agents

GitHub lets enterprises set Copilot agent permissions that users cannot loosen

GitHub made enterprise-managed permissions for Copilot agent operations generally available on Sept. 9. Administrators can block, require approval for, or allow shell commands, file access and network domains, and the company says individual users cannot weaken those limits.

Agents

CrowdStrike announces an identity provider built for AI agents

CrowdStrike on Sept. 2 announced Agentic IdP, an identity provider meant to give each AI agent its own verifiable identity and short-lived access in place of borrowed human credentials. The company's post indicates the product is still in development.

Agents

AWS adds a managed consent portal for agents acting on a user's behalf

Amazon Web Services on Sept. 1 added a managed consent portal to Bedrock AgentCore Identity. It gives users a hosted page for authorizing agents to call outside services for them, replacing OAuth callback plumbing that development teams previously had to build and run themselves.