AI Agents news and analysis on Web Pulse: 37486 curated articles tracking the latest AI Agents developments, tools, and research, updated continuously from vetted sources.
TypeSafe AI introduced a training approach it calls RLCD (reinforcement learning from calibrated decisions), which it says makes models return decisions and probabilities rather than generated text, with outcomes assigne…
Anthropic published a September 2026 snapshot of how it measures frontier AI development, reporting that Claude led 26% of its measured AI R&D work as of August 2026 with more than 90% at least collaborative, and that ro…
The Agentic AI Foundation (AAIF) introduced a sandbox phase in its project lifecycle, announced at its AGNTCon Europe event in Amsterdam, to admit open source agentic AI projects that lack significant adoption, with a 12…
OpenSearch 3.6 and version 2.19 have shipped as the first two releases to carry official Long-Term Support status, a designation the project has lacked since the Elasticsearch fork in 2021. OpenSearch 3.6 also introduces…
OpenAI paid a $6,500 bug bounty to Hacktron AI after three researchers used Anthropic's Claude and OpenAI's Codex to break into OpenAI's GitHub software repository and access a source code component called "Monorepo," th…
GDG Basilicata and the IEEE Student Branch at the University of Basilicata ran an eight-hour student hackathon built entirely on Google's Gemini models via free Google AI Studio API keys, with no local GPU setup. The org…
Palo Alto Networks Unit 42 researchers found that default configurations in AWS AgentCore Harness let prompt injection steer an agent into exfiltrating plaintext credentials managed by AgentCore Identity, because the har…
Enterprise AI deployments lack a lifecycle for managing business context, the meaning attached to data such as how finance defines revenue or which customer policy applies, according to an analysis of enterprise AI conte…
Document AI is shifting from reading individual pages to reasoning across them, according to Forbes Councils member Alberto Gimeno, who describes it as the operating layer that lets autonomous workflows act on what pages…
Google rebuilt its CC AI agent as a shared household assistant that coordinates calendars, tasks, documents, forms, and daily briefings for up to six adults, available as an early Google Labs experiment for US users aged…
Salesforce and Nvidia unveiled Koa, a domain-specific reasoning model built for Salesforce's Agentforce ecosystem, at Salesforce's annual Dreamforce event, where Nvidia CEO Jensen Huang joined Salesforce CEO Marc Benioff…
A September 18, 2026 guide ranked 11 open-source agent harnesses for local LLMs by OSI-approved license, documented local runtimes, maintenance status, and safety controls, with OpenCode taking the top spot for documenti…
Bugsink argues that software products should expose their data through APIs and MCP rather than embedding a separate AI agent in each tool, citing Jira's Rovo, Sentry's Seer, and Salesforce's discontinued Einstein as exa…
Databricks detailed how database branching applies Git-style workflows to databases using copy-on-write, letting developers, CI jobs, and AI agents create isolated environments from a shared database state. In a 40 GB da…
Wontopos, an AI memory company, has launched WOS, a long-term memory API for AI agents that stores end-user memories once and recalls only relevant ones per query to bound LLM input costs. Four tiers are live — Tablet 1,…
Anthropic said on Thursday that its Claude chatbot now leads 26% of the company's research and development work and collaborates with human staff on more than 90% of tasks, warning that models accelerating their own deve…
AI agents are moving from experimentation into everyday business operations, and their ability to access systems, process information, communicate with applications and complete tasks autonomously introduces new security…
A developer argues that AI agents running sequences of tool calls inherit the same distributed-systems problems as event-driven backends, including tail-latency compounding, backpressure, and ambiguous partial failures. …
TypeSafe.ai released Probably 0.1, a small programming language for LLM workflows that is executed by a TypeScript interpreter and powered by Jev for judgments and probabilities. The language uses `feels` for confidence-…
Fulcra Dynamics announced universal multiplayer capabilities on September 18, 2026, enabling agent-to-agent collaboration across any AI model without lock-in, the Boston company said. Fulcra Dynamics, which describes its…