AI Agents news and analysis on Web Pulse: 49187 curated articles tracking the latest AI Agents developments, tools, and research, updated continuously from vetted sources.
Vercel added two new values to the `name` field of its `ListEventType` object: `ai-agent-usage-billing` and `project-git-credential-token-minted`. The change is additive on the producer side but breaks exhaustive switche…
A 2026 analysis of 100,000 paired cold emails found AI-written outreach drew a 4.1% reply rate versus 5.2% for human-written emails, while AI SDR seats sent roughly 7,400 touches per month against 1,150 for human SDRs, a…
Ninety-nine percent of repository access tokens never expire, leaving accounts able to copy or change company code after the work they were granted for ends, according to BlueFlag Security's The State of Developer Risk 2…
Chessigma released version 1.0.0 of its public chess analysis API, a free interface at https://www.chessigma.com/public-api/v1 that requires no API key, signup, or paywall. The API offers four endpoints — /game to submit…
Yuxin Zhu of AntV, the open-source data visualization team behind G2 and G6, published a tutorial for building a command-line data analysis agent using the newly released AVA 4.0 framework, DuckDB, and TypeScript. The ag…
A project called Repairer, published by gno.security, has released an open-source Go program that runs on each server and applies a signed stopgap the day a Linux exploit goes public, then automatically removes it once t…
A new arXiv paper (2611.06969v1) presents an interactive backdoor attack framework that uses training-data poisoning to steer multi-turn LLM-based proactive dialogue systems toward an attacker-specified objective after a…
Researchers built PRISM, an agentic large-language-model pipeline that screened 3.9 million posts across 32 online health communities and surfaced 123 candidate genetic-variant classification differences, 5 of them invol…
A new arXiv paper (2611.07238v1) introduces a unified framework for measuring token waste in AI agents that records and normalizes complete agent trajectories, flags removable spans with an LLM-as-a-Judge, and maps them …
A single-model multi-agent reinforcement learning framework called Octopus, detailed in arXiv paper 2611.06956v1, reaches 49.6% accuracy on BrowseComp-Plus, 17.5 points above single-agent GRPO at 143K context and 12.5 po…
A study posted to arXiv (2611.06953v1) reports that SteelAgent, an LLM-orchestrated system coupling metallurgical descriptors and machine-learning models for steel property prediction, uncertainty reporting and inverse d…
A new arXiv paper (2611.07554v1) finds that persistent memory from benign interactions causes language agents to overgeneralize, a failure the authors call the "memory trap." Across six frontier models, including Claude …
A new arXiv paper (2611.07586v1) introduces TraceMRT, an automated trace-level red-teaming framework that reframes a side task as a coherent sub-goal of an agent's primary task, and reports that black-box attacks built w…
A new arXiv paper (2611.07558v1) introduces SKIM (Selective Knowledge-Informed Masking), which the authors describe as the first pruning and distillation framework for LLM agents operating under structured generation, wh…
A new arXiv paper (2611.06972v1) proposes KDASO, a Knowledge-Data Dual-Driven Automated Skill Optimization method that continuously evolves a skill library for liability adjudication tasks using LLMs. KDASO combines neig…
A new arXiv paper (2611.06962v1) presents AutoIP, a Lean library in which classical linear regularization theory and its underlying functional analysis have been autoformalized by AI agents, with the authors inviting the…
A system called FlexFloor generates approximately 65.40% more response tokens per unit of occupied computing time and raises response generation throughput by 52.74% compared with the verl baseline during reinforcement l…
A new arXiv paper (2611.06995v1) reports that autonomous AI agents systematically diverge from human preferences when formalizing proofs in Lean, based on a "simulated collaboration" analysis of nine major human collabor…
A new arXiv paper (2611.06989v1) proposes the Two-Ledger Model, which joins FOCUS-conformant settlement-grain billing data to an attribution ledger of invocation-grain usage records built on OpenTelemetry GenAI telemetry…
A new arXiv paper (2611.06996v1) empirically characterizes the Trust Responsive Authorization Policy (TRAP), which maps an Authorization Readiness Score combining runtime trust, operational stability and contextual risk …