Do we like them?

Would we support them? Type any of these:

You get the goss with a verdict, a score out of 10, and the sources to check us.

Do we like Google DeepMind?

Alphabet's AI lab behind Gemini and AlphaFold, based in London with a big presence in Mountain View.

Our score: 6 out of 10. Real science that helps people, but it now sits inside a fast-moving Google. Safety transparency, military ties and the Gemini lawsuits stop us backing it fully.

DeepMind was co-founded by Demis Hassabis in 2010 and bought by Google in 2014. It became Alphabet's main AI lab, with Hassabis as its public face. In October 2024 he and John Jumper shared the Nobel Prize in Chemistry for AlphaFold, which predicts protein structures.

The science side has real good in it. DeepMind says over two million researchers have used AlphaFold. It has released predicted structures for over 200 million proteins, and AlphaFold Server is free for non-commercial research. Spin-off Isomorphic Labs is working on drug discovery.

The early days were messy on privacy. The UK data watchdog (the ICO) ruled that the Royal Free hospital trust broke data law when it gave DeepMind 1.6 million patients' records for its Streams app. The ICO did not fine anyone, and DeepMind was the data processor. A later class action was filed in the High Court.

Since then, the lab has become more tied to Google's commercial push. TIME reported that nearly 200 DeepMind staff signed a letter against military contracts. In 2025, 60 UK lawmakers accused it of breaking AI safety pledges over a Gemini release. In August 2026 Hassabis gave up the CEO job amid reported model delays and departures.

Our score is 6. The research is world-class and often useful. The worries are how the safety process is run, military use of the tech, and fresh lawsuits aimed at Google over Gemini. Those lawsuits are allegations, and Google says its safeguards work.

Queue a list of names

One name per line (or split with commas), up to 100 at a time. They're checked a few at a time in the background and land on the scoreboard as they finish. Names already on the board are skipped.

Suggest names

Add names here for the team to review: famous people, characters or companies, one per line (up to 20). We check the ones we pick, and they land on the scoreboard.

Suggested names

Loading…

See what's in line

Loading…

The scoreboard

Everyone checked so far, by score, 10 at the top. Typing a name up top filters the board too. Tap a score to see who's there, and a name for the whole card.

Loading…

Our take, from what the sources we read say on the day we checked. It can miss things; tap a source to read it.