# News: February 20, 2026

> 18 curated links on AI agents, Python, security, Google Cloud, and engineering, selected by Médéric Hurier.

- Author: [Médéric Hurier (Fmind)](https://www.fmind.dev/)
- Published: 2026-02-20
- URL: https://www.fmind.dev/news/2026-02-20/

- [Gemini 3.1 Pro: Announcing our latest Gemini AI model](<https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-1-pro/>): Google launches Gemini 3.1 Pro with enhanced reasoning capabilities and a 77.1% score on the ARC-AGI-2 benchmark, rolling out across the Gemini API, Vertex AI, and consumer products.
- [Gemini 3.1 Pro on Gemini CLI, Gemini Enterprise, and Vertex AI](<https://cloud.google.com/blog/products/ai-machine-learning/gemini-3-1-pro-on-gemini-cli-gemini-enterprise-and-vertex-ai/>): Google Cloud makes Gemini 3.1 Pro available in preview for developers and enterprise customers, offering a more intelligent model for complex, multi-step reasoning tasks.
- [Gemini 3.1 Pro achieves 77.1% on ARC-AGI-2](<https://x.com/arcprize/status/2024522812728496470>): The new Gemini 3.1 Pro model more than doubles the reasoning performance of its predecessor on the ARC-AGI-2 benchmark, setting a new record for general intelligence adaptability.
- [Gemini 3.1 Pro tops the Artificial Analysis Intelligence Index](<https://x.com/ArtificialAnlys/status/2024518545510662602>): Gemini 3.1 Pro has risen to the top of the Artificial Analysis intelligence leaderboard, surpassing Claude Opus 4.6 with an index score of 57.
- [Gemini 3.1 Pro enters the LMSYS Chatbot Arena](<https://x.com/arena/status/2024519891295089063>): The latest Gemini model shows strong competitive performance in the Chatbot Arena, following Gemini 3 Pro's historic achievement as the first model to cross the 1500 Elo threshold.
- [Introducing BigQuery autonomous embedding generation](<https://cloud.google.com/blog/products/data-analytics/introducing-bigquery-autonomous-embedding-generation/>): Google Cloud simplifies AI workflows with a new BigQuery feature that automatically generates and maintains vector embeddings in tables without manual pipelines.
- [Build data agents with Conversational Analytics API](<https://cloud.google.com/blog/products/data-analytics/build-data-agents-with-conversational-analytics-api/>): The new Conversational Analytics API enables developers to build context-aware agents that can query BigQuery data using natural language.
- [How AI is reshaping developer choice (and Octoverse data proves it)](<https://github.blog/ai-and-ml/generative-ai/how-ai-is-reshaping-developer-choice-and-octoverse-data-proves-it/>): GitHub's Octoverse 2025 report shows AI compatibility is a primary driver for developer tool and language adoption, significantly impacting the growth of ecosystems like TypeScript.
- [Turn creative prompts into interactive XR experiences with Gemini](<https://developers.googleblog.com/turn-creative-prompts-into-interactive-xr-experiences-with-gemini/>): Google Developers showcase how Gemini can be utilized to rapidly prototype 3D web graphics and immersive XR simulations from simple text prompts.
- [Monitoring configuration and automating detection & remediation for AI agents](<https://www.youtube.com/watch?v=uaa6VNxcn2s>): A new guide on securing agentic systems using Security Command Center (SCC) to monitor posture and automate threat detection for AI deployments.
- [Use Genie Everywhere with Enterprise OAuth](<https://www.databricks.com/blog/access-genie-everywhere>): Databricks expands access to its natural-language data assistant, Genie, allowing secure integration into Microsoft Teams and custom enterprise applications via OAuth.
- [Mastering Cloud Run costs: The 75% rule for choosing a billing model](<https://www.youtube.com/watch?v=aifIIkQcy4Y>): This technical deep dive introduces the "75% rule" to help developers optimize costs between request-based and instance-based billing on Cloud Run.
- [NotebookLM integrates Gemini 3.1 Pro](<https://x.com/NotebookLM/status/2023851190102986970>): Google's AI research assistant NotebookLM is now powered by Gemini 3.1 Pro, offering users higher limits and improved reasoning for document analysis.
- [Pentagon threatens to sever ties with Anthropic over safety policy](<https://newrepublic.com/post/206631/pentagon-just-sent-terrifying-message-ai-companies>): The Department of Defense is reportedly reconsidering its relationship with Anthropic due to the company's refusal to allow its AI to be used for lethal weapons or warfare.
- [Build AI skills: Google AI Professional Certificate](<https://blog.google/company-news/outreach-and-initiatives/grow-with-google/google-ai-professional-certificate/>): Google launches a new professional certificate to help bridge the AI skills gap, with early adoption from major employers like Walmart and Verizon.
- [Pomelli Photoshoot creates studio-quality marketing assets](<https://blog.google/innovation-and-ai/models-and-research/google-labs/pomelli-photoshoot/>): A new Google Labs tool uses Nano Banana image generation to transform simple product photos into professional-grade marketing assets for small businesses.
- [New productivity features in Chrome](<https://blog.google/products-and-platforms/products/chrome/chrome-productivity-improvements/>): Chrome adds new AI-enhanced productivity features including a multitasking split view and integrated PDF annotation tools.
- [Sovereignty and European competitiveness: A partnership-led approach](<https://cloud.google.com/blog/products/identity-security/sovereignty-and-european-competitiveness-a-partnership-led-approach-to-ai-growth/>): Google Cloud outlines its strategy for AI growth in Europe, focusing on flexible infrastructure that meets regional sovereignty and security requirements.