AI news ·
UK Sovereign AI fund and Stripe CEO invest in Potential Sciences AI drug discovery startup
Potential Sciences raised a seed round from the UK's Sovereign AI fund and Stripe CEO Patrick Collison. The startup claims its physics-based AI can cut chemical design calculation costs by 100 times.
London-based startup Potential Sciences has raised tens of millions of dollars in seed funding from the UK government's Sovereign AI fund and angel investors including Stripe CEO Patrick Collison. The funding, confirmed today, includes a commitment of compute capacity from the Sovereign AI fund - a dual resource injection aimed at accelerating the company's work on AI infrastructure for drug discovery.
Potential Sciences builds AI models grounded in the fundamental laws of physics rather than historical experimental data. The company says this approach could reduce the cost of AI calculations needed to design chemicals by a factor of 100 compared to current methods. Founded in November last year by doctor and former Graphcore executive Hatem Helal, the startup applies machine learning and quantum chemistry simulations to help researchers understand molecular behavior and interactions.
Who is backing the round
The seed round draws from a mix of institutional and individual investors. Venture firms Twin Path Ventures, Odyssey, Portfolio Ventures, Blue Wire Capital, rule30, Nauta Capital, and Lightfield Equity all participated. Patrick Collison joined as an angel investor alongside other unnamed angels. Potential Sciences declined to disclose the exact amount raised or its valuation.
The Sovereign AI fund's involvement goes beyond capital. By providing compute capacity directly, the fund signals a strategic interest in the underlying infrastructure required for computationally intensive drug discovery work. The fund's mandate centers on ensuring the UK develops, owns, and scales key AI companies. Its portfolio already includes chip startup Olix and AI startups Ineffable Intelligence and Callosum.
Founder and technical approach
Hatem Helal founded Potential Sciences in 2025 after his tenure at Graphcore, the chip designer later acquired by SoftBank. His background bridges clinical medicine and semiconductor-scale computing. The company's physics-first modeling strategy departs from conventional AI approaches that rely on training data from prior experiments - a distinction that matters in chemistry, where experimental data can be sparse, noisy, or expensive to generate.
Sovereign AI Investment Partner Joséphine Kant said: "I first met Hatem in Cambridge, and he immediately struck me as a formidable founder. He combines technical excellence with low ego and unapologetic ambition. He knows the science in extraordinary detail, but he's also deliberate about bringing in people who are exceptional in areas different from his own."
Kant added: "Potential Sciences have the ambition to become a true frontier lab for chemistry, with the potential to fundamentally change how the field works."
Why this matters for science and research leaders
The cost claim - 100x cheaper AI calculations for chemical design - is the number to watch. If validated, it alters the economics of early-stage drug discovery, where computational screening costs determine which compounds ever reach a lab bench. For research directors and R&D strategists, physics-grounded models also address a persistent problem: AI systems trained on existing data tend to recapitulate what is already known rather than surfacing genuinely novel molecular candidates. The Sovereign AI fund's compute commitment suggests the government sees this infrastructure as nationally strategic, which may signal future procurement or partnership opportunities for research institutions.