Skip to content
← All news

July 10, 2026

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

  1. OpenAI outlines roadmap for GPT-5 and GPT-6

    openai.com

    OpenAI shared details on the roadmap for its flagship GPT-5 series and the active development of GPT-6, highlighting features like deeper personalization, advanced memory integration, and significantly larger training compute.

  2. Google Cloud Run sandboxes are in public preview

    cloud.google.com

    Google Cloud announced the public preview of Cloud Run sandboxes, a native, secure, and ultra-fast runtime environment built specifically to execute untrusted code and agent workloads starting in milliseconds.

  3. AlphaEvolve is available for everyone

    cloud.google.com

    Google Cloud announced the general availability of AlphaEvolve, a Gemini-powered code optimization and discovery agent designed to solve complex business and computational optimization problems.

  4. LiteRT.js, Google's high performance Web AI Inference

    developers.googleblog.com

    Google introduced LiteRT.js, a high-performance web AI inference runtime that allows JavaScript developers to run machine learning models directly in the browser using WebGPU, WebNN, and WebAssembly fallbacks.

  5. Benchmarking Coding Agents on Databricks’ Multi-Million Line Codebase

    databricks.com

    Databricks shared results from its internal coding benchmark, evaluating AI coding agents on a multi-million line codebase to optimize developer cost and performance.

  6. Overview of AlphaEvolve

    docs.cloud.google.com

    Google released the developer guide for AlphaEvolve, detailing how the specialized AI agent leverages evolutionary heuristics to solve complex combinatorial optimization, mathematical search, and algorithmic discovery problems.

  7. Agentic AI Momentum Report

    aaif.io

    The Agentic AI Foundation published its first Agentic AI Momentum Report, mapping 116 open-source agentic AI projects across five key stack layers: Frameworks & Infrastructure, Agent Capabilities, Agent Applications, Foundations & Protocols, and Operations & Safety.

  8. What’s next in IAM: Security, governance, and runtime defense for AI agents

    youtube.com

    Google Cloud Events hosted a session discussing the future of IAM, focusing on securing, governing, and providing runtime defense for autonomous AI agents.

  9. EU Parliament greenlights Chat Control 1.0 – Breyer: "Our children lose out"

    patrick-breyer.de

    The European Parliament approved Chat Control 1.0, greenlighting the suspicionless mass scanning of private communications despite opposition from a majority of voting Members of the European Parliament.