# News: September 17, 2026

> 9 curated links on AI agents, Python, security, Google Cloud, and engineering, selected by Médéric Hurier.

- Author: [Médéric Hurier (Fmind)](https://www.fmind.dev/)
- Published: 2026-09-17
- URL: https://www.fmind.dev/news/2026-09-17/

- [Inside OpenAI’s agentic software factory](<https://newsletter.pragmaticengineer.com/p/openai-software-factory>): The Pragmatic Engineer explores how Codex powers OpenAI's internal agentic software factory and examines the engineering architecture required to support one billion users.
- [Migrating the GitHub Copilot runtime to Rust, using Copilot](<https://github.blog/ai-and-ml/generative-ai/migrating-the-github-copilot-runtime-to-rust-using-copilot>): GitHub engineers rewrote the Copilot agent runtime into over 800,000 lines of production Rust using Copilot itself across 128 incremental pull requests.
- [Agent Anomaly Detection now in Private Preview on Gemini Enterprise Agent Platform](<https://developers.googleblog.com/agent-anomaly-detection-now-in-private-preview-on-the-gemini-enterprise-agent-platform>): Google launches Agent Anomaly Detection in private preview on the Gemini Enterprise Agent Platform to catch OWASP behavioral risks with zero added runtime latency.
- [TechByte – Gemini Enterprise Product Roadmap](<https://www.youtube.com/watch?v=ctnZAU2162o>): Google Cloud presents the product roadmap for the Gemini Enterprise Agent Platform, detailing secure, cost-effective architectures to build and scale production AI agents.
- [Cloud CISO Perspectives – How Google Monitors AI Threats and Advances AI Defenses](<https://cloud.google.com/blog/products/identity-security/cloud-ciso-perspectives-how-google-monitors-ai-threats-advances-ai-defenses>): Google Cloud's VP of Threat Intelligence Sandra Joyce outlines three structural shifts in the AI threat landscape and explains how Google uses native telemetry to defend AI-assisted pipelines.
- [Mistral x Mozilla](<https://mistral.ai/news/mistral-x-mozilla>): Mistral AI and Mozilla partner to integrate Mistral models into Firefox's Smart Window assistant, delivering privacy-focused, zero-data-retention AI capabilities directly into the browser.
- [AI researchers debate how close we are to recursive self-improvement](<https://www.youtube.com/watch?v=PrSf7IOYu-I>): Dwarkesh Patel hosts AI researchers John Schulman, Charlie O’Neill, and Beren Millidge to debate the timelines, empirical evidence, and bottlenecks of recursive self-improvement in frontier models.
- [Gmail Search’s AI Overviews now available globally](<https://workspaceupdates.googleblog.com/2026/09/gmail-searchs-ai-overviews-now-available-globally.html>): Google rolls out AI Overviews in Gmail search globally to paid Workspace and Google One AI Premium users with English language settings, enabling natural language inbox summaries.
- [PS5 Linux lead quits as open-source projects have become “a bunch of noobs using LLMs” that “they don’t even understand”](<https://frvr.com/blog/news/ps5-linux-lead-quits-as-open-source-projects-have-become-a-bunch-of-noobs-using-llms-that-they-dont-even-understand>): The lead developer of the ongoing PS5 Linux project leaves the homebrew scene after open-source development was overrun by low-effort LLM code contributions and AI vibe-coders.
