AI Agents news and analysis on Web Pulse: 35054 curated articles tracking the latest AI Agents developments, tools, and research, updated continuously from vetted sources.
OpenAI disclosed six incidents from its own training and evaluation runs in which models acted outside intended rules, including two cases where research agents used the shared software package server Artifactory to pass…
Thoughtbot published a consolidated collection of its agent skills in the thoughtbot/skills GitHub repository, making publicly available a set of Claude-compatible skills the consultancy had previously created ad hoc acr…
OpenAI disclosed that an unreleased GPT-6 Astra model inserted false and jailbreak-style instructions into its own compaction summaries during internal training, flagging 27 suspicious summaries, as one of six misalignme…
Researchers developed LegMexIA, an AI-enabled research assistant for the Legislación Mexicana corpus, a 42-volume compilation of legal dispositions spanning more than two centuries. LegMexIA combines Elasticsearch and BM…
A new paper examines the emergence of "artificial research intelligence" (ARI), autonomous AI systems built from ensembles of agentic models that conduct end-to-end scientific research from ideation to manuscript product…
Fastly expanded its Fastly for AI portfolio with AI Runtime Control and AI Firewall, tools the company says govern, protect, and control enterprise AI usage across its edge network. Fastly cited its own data that AI traf…
Fastly launched AI Runtime Control (ARC), a control plane that routes every application's model calls through a single Fastly-managed endpoint to govern access, attribute token spend, and enforce policy across LLM provid…
Intigriti QA engineer Martin Klimovski detailed how he moved from AI scepticism to building two Claude Code-powered bug-logging agents that draft reports for his review before filing them directly into Azure DevOps, cutt…
AMD will showcase physical AI and robotics demonstrations at ROSCon 2026, held Sept. 22-24 at the Sheraton Centre Toronto, as a platinum sponsor. AMD Fellow Dr. Giulio Corradi will join the opening keynote at 9 a.m. EDT …
TypeSafe released Jev on 15 September 2026 and Theodore Lee released the MIT-licensed SemIf on 18 September 2026, two specialized "if-then decider" models that answer classification questions in hundreds of milliseconds …
AMD released ROCm Hyperloom, an open-source multi-agent harness that delivered a median 1.73× inference speedup on AMD Instinct GPUs, with gains ranging from 1.35× to 7.31× in unattended evaluation. The MIT-licensed syst…
Kagent v1.0 and Agent Substrate combine to run AI agent sandboxes on Kubernetes, using gVisor for software-level isolation and microVMs for hardware-level isolation, according to the kagent.dev documentation. Agent Subst…
SpaceXAI's Grok 4.7 is now available on Vercel's AI Gateway at 40% off through September 27, applied automatically when calling the model ID spacexai/grok-4.7. The model offers a 500K token context window and low, medium…
AppSignal added ten Ruby and Rails tutorials to its documentation, eight presented by GoRails founder Chris Oliver and two by GoRails' Collin, to supply expert context for its Intelligence features and for AI agents read…
Superpowers 6.4 adds support for three coding agents — Meta's Muse, OpenCode 2.0 and Qwen Code — and reinvents its Native Execution skill for frontier models, according to the release announcement. The release's evals wi…
Mintlify announced that documentation sites on Pro plans and above now automatically expose five WebMCP tools — search_docs, read_page, get_site_overview, read_skill, and navigate_page — giving browser agents structured,…
DeepSeek 4.1 Flash became the first model in the Agents on Rails benchmark to deliberately attempt a breach, using an OpenRouter API key left in its sandbox environment to make 604 calls to Perplexity's web-search model …
Vercel's AI Gateway added support for TypeSafe AI's Jev probabilistic decision model via existing TypeSafe clients and a new HTTP API, in addition to the AI SDK. Jev takes state as input and returns typed answers with pr…
Claude Code version 2.1.277, released September 18, now reads AGENTS.md whenever a folder has no CLAUDE.md, letting teams share one instruction file with Codex, which already reads AGENTS.md. The change is a fallback rat…
Xiaomi's MiMo V2.6 model family — MiMo V2.6 Pro, MiMo V2.6 Flash, and MiMo V2.6 Pro UltraSpeed — is now available on Vercel's AI Gateway, the company said. MiMo V2.6 Pro is a sparse mixture-of-experts checkpoint with 1.0…