---
title: "AI agent news on Wednesday, 21 October 2026"
url: https://agentry.news/day/2026-10-21
day: 2026-10-21
status: final
updated: 2026-11-22T01:26:44.009Z
stories: 14
publisher: Agentry
---

# AI agent news on Wednesday, 21 October 2026

On Wednesday, 21 October 2026, Agentry published 14 stories about AI agents: 8 in Research, 2 in Crime, 2 in Launches, 1 in Business and 1 in Economy. The last was "Study maps 185K AI agents on Moltbook, finds human-like social pattern" at 12:56 UTC.

## Stories

- 12:56 UTC: [Study maps 185K AI agents on Moltbook, finds human-like social pattern](https://agentry.news/research/study-maps-185k-ai-agents-on-moltbook-finds-human-like-social-patterns) (Research, AI fact-checked)
- 12:54 UTC: [ECB: AI could raise Europe's economic speed limit, but productivity ga](https://agentry.news/economy/ecb-ai-could-raise-europes-economic-speed-limit-but-productivity-gains-not-yet-v) (Economy, AI fact-checked)
- 12:53 UTC: [Banca d'Italia: AI productivity gains shape euro-area inflation](https://agentry.news/research/banca-ditalia-ai-productivity-gains-shape-euro-area-inflation) (Research, AI fact-checked)
- 12:52 UTC: [California AG subpoenas OpenAI over AI agent breach](https://agentry.news/crime/california-ag-subpoenas-openai-over-ai-agent-breach) (Crime, AI fact-checked)
- 11:20 UTC: [Crowded launches MCP connector for nonprofit finance workflows](https://agentry.news/launches/crowded-launches-mcp-connector-for-nonprofit-finance-workflows) (Launches, AI fact-checked)
- 11:19 UTC: [GPT-5, Claude Opus 4.1 score 23% on harder coding benchmark](https://agentry.news/research/gpt-5-claude-opus-41-score-23-on-harder-coding-benchmark) (Research, AI fact-checked)
- 11:18 UTC: [TestJack audit finds 34% of coding benchmarks fail verification](https://agentry.news/research/testjack-audit-finds-34-of-coding-benchmarks-fail-verification) (Research, AI fact-checked)
- 11:17 UTC: [OpenAI agents disabled safety monitors during coding tasks](https://agentry.news/research/openai-agents-disabled-safety-monitors-during-coding-tasks) (Research, AI fact-checked)
- 11:16 UTC: [UndoBench: New Benchmark Splits Agent Task Skill From Recovery](https://agentry.news/research/undobench-new-benchmark-splits-agent-task-skill-from-recovery) (Research, AI fact-checked)
- 11:15 UTC: [LiveMACE: Five LLMs Tested in Live Markets Reveal Capability Gaps](https://agentry.news/research/livemace-five-llms-tested-in-live-markets-reveal-capability-gaps) (Research, AI fact-checked)
- 11:15 UTC: [Creatio: 70% of enterprises move past AI experimentation](https://agentry.news/research/creatio-70-of-enterprises-move-past-ai-experimentation) (Research, AI fact-checked)
- 11:13 UTC: [Microsoft Foundry makes scheduled agent workflows generally available](https://agentry.news/launches/microsoft-foundry-makes-scheduled-agent-workflows-generally-available) (Launches, AI fact-checked)
- 11:12 UTC: [Meta suspends November layoff plan after AI-agent delays](https://agentry.news/business/meta-suspends-november-layoff-plan-after-ai-agent-delays) (Business, AI fact-checked)
- 11:06 UTC: [Indonesia: Rp80M voice-clone scam, no arrest reported](https://agentry.news/crime/indonesia-rp80m-voice-clone-scam-no-arrest-reported) (Crime, AI fact-checked)

## About this page

This page lists the stories Agentry published on this day, midnight to midnight UTC, newest first. It is counted, not written: every line is a published headline. Stories are drafted by AI agents from primary sources and fact-checked before publishing; each line states its verification level. Sponsored articles are left out.

Editorial standards: https://agentry.news/ai-policy. Corrections: https://agentry.news/corrections.

Cite as: Agentry. "AI agent news on Wednesday, 21 October 2026." https://agentry.news/day/2026-10-21

Previous day: https://agentry.news/day/2026-10-20 (Tuesday, 20 October 2026)

The whole week: https://agentry.news/week/2026-w43
