Google DeepMind Releases Suite of Advanced AI Systems

Google DeepMind has published three significant AI systems across genomics, weather forecasting, and robotics, marking a major acceleration in applied AI research.

Genomics and Weather Frontiers

On September 1, 2026, Google DeepMind published AlphaGenome Atlas, described as a high-resolution map of human DNA. The same date saw Google DeepMind and Google Research jointly publish WeatherNext 3, their most advanced and accurate global weather AI model.

Robotics Advances

Google DeepMind has made steady progress in embodied AI. In July 2026, the company published Gemini Robotics 2, described as bringing whole-body intelligence to robots. Earlier, in April 2026, Google DeepMind published Gemini Robotics ER 1.6, which powers real-world robotics tasks through enhanced embodied reasoning.

Multi-Agent Research Collaboration

Two joint initiatives launched in May 2026 represent a shift toward AI-assisted scientific discovery. Google DeepMind, Google Cloud, Google Labs, and Google Research jointly published Co-Scientist, described as a multi-agent AI partner to accelerate research. On the same date, Google Labs, Google DeepMind, and Google Research jointly published Gemini for Science, described as AI experiments and tools for a new era of discovery.

OpenAI Reaches Automated Research Intern Milestone

OpenAI stated it reached its goal of having an automated ‘research intern’ by September 2026—a system able to handle well-defined tasks that would otherwise take a skilled researcher several days.

The achievement reflects unprecedented scaling in AI-accelerated research. By mid-August 2026, the median OpenAI researcher was spending more than $600 a day on coding-agent inference at API prices, with the top 10% of researchers spending more than $7,000 in tokens daily on coding-agent inference.

OpenAI’s research organization logs 3.1 agent-workdays of effort for every single human workday, a threshold crossed sometime before June 2026. Every category of OpenAI research activity—including writing code, running training evaluations, and monitoring live experiments—grew between January and August 2026.

Security Incident

A security breach involving OpenAI’s own coding agents forced researchers to pause parts of their AI-accelerated research work over the summer of 2026.

Source: Google AI