Who is trying to build superintelligence?
A few years ago, saying you wanted to build superintelligence sounded like science fiction. Now several of the world’s largest companies say it openly.
Published 6 October 2026 by Omar Nouha Sane
This overview reflects the situation in autumn 2026. Companies, leaders and strategies in this field change quickly.
The main labs
- OpenAI
- Founded in 2015 with the mission of making sure AGI benefits all of humanity. CEO Sam Altman wrote in early 2025 that the company was confident it knew how to build AGI and was turning its attention to superintelligence.
- Google DeepMind
- Google’s combined AI lab, led by Demis Hassabis, who shared the 2024 Nobel Prize in Chemistry for protein structure prediction. DeepMind has described AGI as its long-term goal since its founding in 2010.
- Anthropic
- Founded in 2021 by former OpenAI researchers, including CEO Dario Amodei. It describes itself as focused on AI safety research while building frontier models, and has published its expectation that very powerful AI may arrive within a few years.
- Meta
- In 2025 Meta created Meta Superintelligence Labs and hired researchers from rival labs with very large pay packages, making superintelligence an explicit company goal.
- xAI
- Elon Musk’s AI company, founded in 2023, which develops the Grok models. Musk has repeatedly given some of the earliest timelines for AI surpassing humans.
- Safe Superintelligence Inc. (SSI)
- Founded in 2024 by Ilya Sutskever, OpenAI’s former chief scientist, together with Daniel Gross and Daniel Levy. SSI says it has one goal and one product: a safe superintelligence. Sutskever became CEO in 2025.
Beyond the US
Chinese labs such as DeepSeek and Alibaba’s Qwen team have released highly capable models, often with openly available weights. Their progress is one reason US policy increasingly frames AI as a race between countries.
Governments
Governments are not building superintelligence themselves, but they shape the race through chip export rules, funding and regulation. In September 2026 the White House went as far as renaming AI to “super intelligence” in official federal language. Our guide on the SI order explains what that does and does not change.
Why it matters for the forecast
The number of well-funded groups working on the same goal is one reason forecasts have moved earlier. It also raises the stakes for alignment: when many players compete, there is pressure to move fast, and safety work can fall behind. Some markets on the tracker even ask which lab will get there first.