This Week in AI

The Editor

20 Feb 2026 — 2 min read

AI this week was defined by scale and risk: massive funding moves, faster specialised models, and platform-wide upgrades from major players like Google and OpenAI. At the same time, real-world incidents — outages, security bans, and governance debates — showed how quickly agentic systems are colliding with enterprise reality.

Google ships Gemini 3.1 Pro (and pushes it across Gemini + Vertex AI)

Google announced Gemini 3.1 Pro as a smarter baseline for complex problem-solving, rolling it out across the Gemini app, APIs, and Vertex AI. The real story isn’t just benchmarks — it’s Google treating model upgrades as platform-wide infrastructure changes that hit consumers, developers, and enterprise simultaneously.

OpenAI mega-fundraising reportedly approaches $100B+

Reports suggest OpenAI is finalising a massive funding round potentially exceeding $100B, pushing valuations toward the ~$850B range. The scale highlights how AI infrastructure is now competing with sovereign-scale capital projects, with strategic investors focusing on compute and model ecosystems rather than traditional SaaS metrics.

OpenAI launches GPT-5.3-Codex-Spark for real-time coding

OpenAI introduced GPT-5.3-Codex-Spark, a low-latency coding model optimised for real-time development workflows. Built alongside Cerebras hardware, the release signals a shift toward specialised inference stacks designed for interactive engineering tasks rather than pure reasoning benchmarks.

AI tooling triggers real-world cloud reliability incident

Reports describe outages linked to AI-powered coding tooling interacting with AWS infrastructure. While framed as “user error,” the event highlights a growing risk surface: agentic automation operating close to production environments without traditional safety rails.

OpenClaw security fears push companies to restrict agentic tooling

The viral agentic platform OpenClaw is facing bans and restrictions from major tech firms after security vulnerabilities and unpredictable behaviour surfaced. Combined with GitHub advisories, the situation highlights a broader industry trend: fast-moving agent ecosystems colliding with enterprise governance expectations.