Google DeepMind

Alphabet's AI lab (2023 merger of DeepMind and Brain); builds the Gemini models.

London, UK · 2010

Analysis as of Sep 25, 2026

In one week, Washington narrowed who may test frontier models first — and the labs drafted rules for who may audit them

The White House asked OpenAI and Anthropic to keep new frontier models from the UK AI Security Institute until a US review is finished, Politico reported on September 24. The same week Google, OpenAI and Anthropic were reported to be close to launching a self-regulatory standards body that would set qualifications for independent auditors. Both decisions are about who checks the labs. Neither was made by an independent checker.

Our read — labelled opinion, not investment advice.

Anthropic

OpenAI

Google DeepMind

Press · 2026-09-24 Read

Largest training compute — 1×10²⁶ FLOP, measured series

Explainer Sep 6, 2026

Frontier AI: where the field actually stands

Training compute passed 1e26 FLOP and grows 4–5× a year, and in 2026 a frontier model was export-controlled like a strategic technology for the first time. The benchmarks keep moving: ARC-AGI-3 launched with every model under 1%, then one reached 30.2%.

Read

Milestone note Jul 30, 2026

Gemini Robotics 2 gives humanoids whole-body control — demonstrated on Apptronik's Apollo 2

Google DeepMind released Gemini Robotics 2, extending robot control from upper-body manipulation to whole-body coordination — 'feet to fingertips.' Demonstrated on Apptronik's Apollo 2, it walks, crouches and manipulates while reasoning through instructions in real time. Dexterity results include a 92% success rate unscrewing a lightbulb; a companion ER 2 model handles planning and multi-robot teamwork.

Google DeepMind

Apptronik

Press · 2026-07-30 Read