The Council for Science and Technology (CST) has issued detailed recommendations on how the UK could establish a sovereign artificial intelligence (AI) chip design industry, identifying it as a once-in-two-decades opportunity.
With global demand for AI chips forecast to grow by 30% per year until 2030 – reaching more than half of total semiconductor revenues – the UK faces both an opportunity and a challenge: how to secure a profitable stake in this market.
A strategic focus on AI chip design
The CST emphasised that the UK should not attempt to replicate large-scale semiconductor manufacturing, given the prohibitive costs and established global leaders. Instead, it advised concentrating on AI chip design, where the UK’s strengths in research, creativity, and AI expertise can be leveraged.
Chip design, compared to fabrication, is less capital intensive and more compatible with a ‘fabless’ model, where designs are produced domestically but manufactured by overseas foundries such as TSMC or Samsung.
The UK’s opportunity
Despite a decline in the national semiconductor industry since the successes of ARM and Inmos, the UK retains expertise in compound semiconductors, optoelectronics, and power electronics. This foundation, combined with world-class universities and research centres, could support a new wave of innovation in AI chip design.
The CST warned, however, that without a sovereign design capability, the UK risks over-reliance on a single global supplier, threatening supply chain resilience for both commercial and defence applications.
To establish a profitable industry, CST set an ambitious objective: enabling UK companies and overseas design centres with UK operations to launch 50 new AI chip products within five years. Achieving this would require government-led coordination across skills, investment, and infrastructure.
Key recommendations
Skills
Expand chip design education: The Department for Science, Innovation and Technology (DSIT) and the Department for Education (DfE) should fund bursaries and fellowships to train 12,000 new chip designers by 2030.
National chip design curriculum: universities should reintroduce comprehensive chip design courses, supported by leading institutions such as Edinburgh, Bristol, Southampton, and Sheffield.
Optoelectronics training: increased investment in the Optoelectronics Research Centre would prepare engineers for a future where AI chips rely on optical interconnects.
Investment and finance
Strategic objectives: DSIT and the Ministry of Defence (MOD) should jointly set priorities for semiconductor investment, including dual-use opportunities.
Innovation pipeline: investment should cover the entire pathway, from early-stage Innovate UK grants to British Business Bank support for scaling. The National Wealth Fund should fund infrastructure, such as advanced chip packaging.
Prioritisation: the government may need to deprioritise funding in slower-growth semiconductor segments, focusing instead on AI chips and optoelectronics.
Infrastructure
Affordable facilities: the UK Semiconductor Centre should ensure small firms and academics can access design and testing infrastructure cost-effectively and quickly.
Access to global technologies: the UK should negotiate national agreements with foundries and EDA tool providers to secure access for startups, potentially through trade agreements or national-level licensing.
Looking ahead
The CST report concluded that creating a sovereign AI chip design industry is both an economic and strategic imperative. Without bold government commitments, the UK risks missing a significant growth opportunity and becoming dependent on foreign suppliers. With them, however, the UK could establish itself as a global player in one of the fastest-growing technology markets.
