AI Will Cure Cancer, Says Arm's CEO — If the Chip Supply Holds

In an interview that oscillates between the visionary and the sobering, Rene Haas — chief executive of Cambridge-based chip designer Arm Holdings — has told the BBC that artificial intelligence will find a cure for cancer that humans cannot, within our lifetimes. He paired that prediction with a second, equally bold one: AI-driven humanoid robots will be in widespread use within five years. And then, having handed out two optimistic forecasts, he delivered the caveat that arguably matters most to both of them: the industry cannot get enough chips.

The comments came in the BBC’s “Big Boss Interview” podcast with economics editor Faisal Islam, published on September 8, 2026, and quickly picked up by the Guardian, Business Insider, and outlets across Europe and Asia. For a company that rarely dominates headlines the way Nvidia or OpenAI do, the interview is a reminder of just how central Arm has become to the AI economy — and how its chief executive sees the next decade unfolding.

“Too complex a problem” — for now

Haas, who stepped down from the board of British pharmaceutical giant AstraZeneca in April, was specific about why cancer has resisted conventional approaches. “Modelling a cell, modelling a human, modelling how a DNA marker is impacted by cancer — it’s too complex a problem, not only for humans today, but the computers that run AI,” he said. “However, going forward, as we feed more and more of the models into these computers, and the computers get more sophisticated to run the models, they’re going to solve it.”

His phrasing is careful. This is not a claim that today’s models are on the verge of an oncology breakthrough. It is a scaling argument: as compute grows and models ingest more biological data, problems that are currently intractable become solvable. “AI is going to… find a cure for cancer that today you and I, other humans [could] not in our lifetimes,” he said. “I believe in our lifetime, AI will help cure cancer.”

Not everyone in the field frames it quite that way. Prof. Chris Bakal of the Institute of Cancer Research, London, and CEO of Sentinel4D, told the BBC that the real question is no longer whether we use AI, but what we feed it. In labs like his, models are trained on data generated from patient samples — not scraped from the internet, and not dependent on a giant data centre to run. “The future of medical AI will not belong to whoever builds the biggest computer,” Bakal said. “It will belong to whoever has the right measurements.” His counterpoint is a genuine tension in the field: brute-force scaling versus high-quality, purpose-built biological datasets. Both camps believe AI will cut years off drug development; they differ on where the binding constraint lies.

Robots in five years, and a chip shortage standing in the way

On robotics, Haas was similarly aggressive. He said the chips Arm designs will fuel self-learning AI robots “in a very large way” in manufacturing, cleaning, security, bridge construction, and repairs within a decade. The near-term milestone — widespread humanoid robots within five years — is a timeline that would have sounded outlandish two years ago but now sits comfortably inside the range that major lab leaders and robotics firms are publicly discussing.

“With artificial intelligence, these robots can see, learn, and essentially be reprogrammed for new tasks,” he said. “The robot that was programmed to make a bed can also learn how to arrange the towels in a room, or clean the dustbins.”

On jobs, Haas pushed back against the more alarming forecasts: “the estimates of jobs going away and being completely replaced by machines is a bit overstated,” he said, arguing that new opportunities will outweigh displacement. He applied similar reasoning to fears of an AI stock correction, citing long-term demand — a reassuring line, though one delivered by an executive whose company’s share price has been a direct beneficiary of that boom. Earlier this summer, Arm’s peak valuation made it, in cash terms, the most valuable UK-based company in history.

The supply-constrained decade

The most concrete part of the interview concerned supply. Arm’s power-efficient technology is now used in roughly half of AI data centres worldwide, Haas said — a striking figure for a company long associated with phone and embedded chips. That position has driven a significant evolution in Arm’s business: the company now sells its own microchips, and Facebook-owner Meta asked it to develop the Arm AGI chip. Demand, in Haas’s words, has been “off the charts” — more than $2 billion worth since the chip launched in March.

But supply has not kept pace. “We are absolutely in a supply-constrained environment,” Haas said, pointing to planned multi-gigawatt data centres in France and the US, and even proposals to put data centres in space. “Can you get enough chips? We need more fabs before we put a data centre in space.”

His candour extended to the UK’s own industrial ambitions. Despite government talks about bringing parts of the chip supply chain to Britain, Haas was blunt: “I don’t think it’s necessary for the UK to [build] fabs. Fabs are very expensive. They take a lot of specialised workers, they take a lot of natural resources, and there’s a pretty broad ecosystem for those.” He noted that half of Arm’s employees remain in the UK and that the company is “by far and away the largest employer in Cambridge” — a defence of Arm’s 2016 sale to SoftBank and its subsequent Nasdaq flotation, decisions that some British ministers still lament.

Why this matters

Strip away the headline grab and the interview is a fairly complete statement of the compute-optimist worldview: scaling solves science, robots follow, jobs adapt, and the only real enemy is fabrication capacity. It is also a convenient worldview for the company that licenses the designs inside a growing share of the world’s AI infrastructure.

What gives it weight is Arm’s unusual vantage point. The company sits inside SoftBank — which also holds major AI investments, including in OpenAI — and sees demand signals from hyperscalers, chipmakers, and device makers simultaneously. When its CEO says supply, not science, is the binding constraint on AI-driven drug discovery, that is a data point from the supply chain itself, not a lab’s press release.

Whether AI “cures cancer” in our lifetimes is a claim no one can verify today. But the underlying wager — that compute growth converts into biological insight — is already being tested in oncology labs, and the $2 billion of demand for a single Arm chip design suggests the market believes the compute side of that bet will be funded either way. The bottleneck, if Haas is right, will be made of silicon.