Imagine a laboratory that can use the result of one experiment to suggest the next useful test. That is the idea behind King's Autonomous Labs, a new research initiative announced by King's College London on 28 September.
From one experiment to the next
King's is bringing together AI, robotics, sensors and laboratory automation to explore what it calls self-driving laboratories. The proposed loop is straightforward to describe: run an experiment, analyse the results, work out what should be tested next, then run the next experiment. As new evidence comes in, the next test can be adjusted rather than following an unchanging sequence.
Sensors and laboratory equipment collect observations. AI can help interpret those observations and suggest a useful next step. Robotics and automation could then carry out a further test within limits set by researchers. The goal is to make experimentation more responsive to what scientists actually find.
Self-driving does not mean scientist-free. Scientists remain responsible for the questions, objectives, boundaries and safeguards, including when a result needs a person to make the call. King's says the initiative is intended to extend researchers' expertise, not replace it.
Sources: King's College London: launch announcement, 28 September 2026 ↗ · King's Autonomous Labs: programme and funding ↗
A small start with living systems
King's has committed £150,000 to an initial funding programme for rapid, six-month proof-of-concept projects involving living systems. Its first challenge includes research on cells, tissues, organoids and microbes. Living systems can change unexpectedly, making them a demanding test for any system that must decide how to respond to new measurements.
The early projects are intended to develop pilot data, demonstrators and ways of working. King's hopes to learn what these systems can do, where human judgement is essential and how to build appropriate safeguards before pursuing wider applications.
Read more: King's Autonomous Labs: living systems and pump-priming funding ↗
Why the distinction matters
The practical promise is a tighter cycle between testing and learning. But launching a programme is different from demonstrating that it works at scale. For now, this is research into what responsible laboratory autonomy could look like, not evidence of faster medicines or new treatments.
The useful question is not whether a laboratory can run without scientists. It is whether well-designed tools can help scientists test ideas more effectively while keeping people accountable for the work.
No AI for the sake of AI.
Less hype. More real-world AI progress.