Insilico Medicine says its AI-assisted drug candidate ISM9528 performed better than morphine in preclinical models. The potential non-opioid treatment is due to move into human testing, but it has not yet shown that it works or is safe for people.
Watch Desk analysis
What happened
The candidate is intended to treat acute and chronic pain. Insilico’s president and co-founder, Alex Aliper, told The National that the company sees it as a possible alternative to opioid medication. He described the preclinical results as unprecedented; that is the company’s characterisation, not evidence of a proven treatment in patients.
The company says its Pharma.AI platform helps researchers identify disease-related biological mechanisms, select targets and design molecules for laboratory testing. Insilico estimates that its process can take 12 to 18 months to nominate a preclinical candidate, compared with about five years in traditional drug discovery. Those are company estimates for early-stage discovery, not a promise that a medicine reaches patients sooner.
Why it matters
A potential non-opioid pain medicine is a meaningful target, and this report gives a concrete example of AI being used in drug discovery: helping select biological targets and design candidate molecules, which are then synthesised and tested. The next step towards human trials makes ISM9528 more than a computational proposal, but it remains an early candidate. Insilico says AI does not shorten the clinical-trial process, which can take about a decade and remains subject to regulatory review.
Our read
The interesting development is not that an algorithm has solved pain. It is that an AI-discovered candidate has reached the point where human testing is being discussed, with preclinical results still a long way from a treatment people can use. Promising is welcome; patients deserve the evidence one careful trial at a time.
What to watch
- Whether ISM9528 enters human trials, and what those studies are designed to test.
- Whether clinical results support the reported preclinical comparison with morphine.
- What evidence emerges about safety, effectiveness and the risk of dependence.
Discussion spark: Should AI-discovered drug candidates be judged by how quickly they reach human trials, or only by the clinical evidence they eventually produce?
Sources and evidence
- Source update (16 October 2026, 12:52 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.