Event round-up: Making the case for semiconductors: The role of chips as a foundation for AI and Deep Tech

Charlie opened by setting out the policy landscape within which the UK's semiconductor work is currently situated, before turning to the structure of his team and the mechanisms being developed for cross-sector coordination. 

At the foundation is the Industrial Strategy, which designates eight high-growth sectors. For Charlie's team, the most directly relevant is the Digital & Technologies sector plan, owned by DSIT and intended as a framework for long-term strategic direction. Semiconductors sit within this as one of six designated frontier technologies — alongside advanced connected technologies, AI, cybersecurity, engineering biology, and quantum. Charlie stressed the importance of semiconductors being included in this plan as a foundation for making the case for continued government support, and as a means of unlocking cross-technology collaboration across the other frontier tech teams. 

One headline ambition from the D&T sector plan is securing the UK's first trillion-dollar tech company — a goal Charlie noted could plausibly be achieved through existing semiconductor companies. More broadly, the plan aims to crowd in greater private sector investment, which Charlie described as a central objective for the current government. 

On the structure of his team, Charlie outlined efforts to consolidate related technology areas — including robotics and advanced materials — under a single directorate, enabling both practical day-to-day collaboration and more joined-up strategic decision-making from the top. 

Charlie also highlighted emerging coordination mechanisms within UKRI, including the establishment of programme boards designed to bring together UKRI and government leadership to provide strategic oversight of programme design, delivery, and performance. A dedicated semiconductor programme board is being stood up, alongside an anticipated overarching frontier tech board to ensure alignment across technologies. 

He closed by pointing to the UK Semiconductor Centre as a key vehicle for the ecosystem coordination that events like this one are designed to support. Two of the Centre's five missions — ecosystem & advocacy, and strategy & roadmapping — are specifically aimed at creating a focal point for industry, academia, and government. He encouraged attendees to engage with the Centre, with techUK, and with DSIT directly, to help shape priorities and surface barriers. 

George introduced the British Business Bank as a 100% government-owned economic development bank operating at arm's length — given a mission and mandate by government, but making independent investment decisions. His own focus within the bank sits in the direct investments team, which he described as overweight on frontier technologies like quantum and semiconductors relative to the broader venture capital market — a positioning he attributed less to prescience and more to an early recognition of a significant and unwarranted market gap in R&D-intensive companies. 

A core theme of George's presentation was the persistent financing gap between UK businesses and their US equivalents. Drawing on analysis published in the bank's Small Business Equity Tracker, he highlighted that UK companies consistently receive less funding than comparable US businesses at every stage — a gap that becomes significantly more pronounced at later funding rounds, and is particularly acute for deep tech and life sciences. He noted that the comparison likely understates the true picture, given that it measures the whole of the US rather than the concentrated clusters of the West Coast or East Coast. 

Despite the UK being the third-largest VC market in the world, this gap remains meaningful when benchmarked against the outlier of the US — and addressing it sits at the heart of the British Business Bank's mandate to drive economic growth through interventions in access to finance. 

George also noted that the bank now has more capital to deploy from government than in previous cycles, with a larger allocation directed toward direct investments — meaning its new investments are larger than has historically been the case. 

•  Aubree Zimmer, Director of the BEDI Project, CORNERSTONE  
•  Cian Bates, Tech First Programme Lead, DSIT 

Aubree opened the session with a prop — a handful of miniature bricks, sourced from the Brickworks in Swanwick, just outside Southampton: the last surviving steam-driven brickworks in the south of England, and one that at its peak produced over 20 million bricks. The analogy was deliberate. Bricks, she explained, were once a foundational technology — standardising construction, enabling scale, underpinning the building of railways, ports, factories and homes. Semiconductors, she argued, occupy that same role today: not merely a frontier technology, but a foundational one. The challenge now is painting that picture for a wide range of audiences. A phrase she returns to often, and hears frequently from students, framed the session: this isn't for people like me. 

The domestic talent pool: strengths and gaps 

Cian leads the Tech First programme within DSIT. He described Tech First as an effort to bring frontier industries to life — showing people how technology is made, developed, broken, and fixed — in the same spirit as Aubree's bricks. 

On the state of the domestic talent pool, Cian was direct. The UK has strong foundations — world-class universities and further education institutions — but two significant challenges stand out. The first is diversity. Around 26% of the semiconductor workforce is female; in cybersecurity, the figure drops to 18%. Attracting more girls and women into the sector, at both entry and mid-career levels, is a priority. The second is pathways — making it clear what a career in each of the six frontier industries looks like, what skills are needed at every stage, and how to progress. 

Aubree built on both points. At CORNERSTONE, she has developed the BEDI Project — a programme dedicating £5,000 over five years and 10% of contracted time per eligible participant to support career progression through courses, events, and professional memberships. Findings from this work, funded by EPSRC, are due to be published in five years. But she was keen to push beyond entry points. Citing the Lovelace Report (2025), she noted that the lack of retention for women is currently estimated to cost the UK economy £2–3.5 billion a year. Getting people in is only part of the problem; keeping them is the other. 

Initiatives targeting retention and mid-career women 

Cian outlined two government-led responses. The first is the Secretary of State's Women in Tech Task Force, specifically examining why retention rates for women in digital and tech are low and how to address that sustainably. The second, under the Tech First banner, is the recently launched Tech First Women's Programme — targeting mid-career women and supporting their entry or re-entry into digital and tech sectors, with a particular focus on routes into SMEs. 

Where meaningful engagement starts: lessons from Cyber First 

On the question of when engagement needs to begin, Cian pointed to the track record of Cyber First — Tech First's predecessor programme, now a decade old — as the clearest evidence base. When Cyber First launched its scholarship in 2015, around 19% of applicants were women. The National Cyber Security Centre responded by building a pipeline starting at age 11–13, before GCSE choices, with the aim of exciting young people about what technology is and what it can do — not turning them into experts, but sparking curiosity. The Cyber First Girls Competition, which most recently reached 10,000 twelve- and thirteen-year-olds, has become a flagship example. Most recently, nearly 40% of scholarship applicants were women, many of whom cited early engagement with the pipeline as the reason. Tech First is built on replicating and expanding that model across all six frontier industries. 

Reaching the right audiences, with the right message 

Aubree pressed on a challenge she sees consistently: the gap between building a programme and actually reaching the people it's designed to serve. Cian acknowledged this, noting that Tech First has had to engage not just students, but teachers — thinking about how semiconductor careers connect to the computing and physics curriculum — and parents, who need to understand the range of specialisms available in digital and tech in the same way they might understand law or medicine. A practical frustration he highlighted: if someone searches today for "how do I become a semiconductor professional," the first ten results give ten different answers. Simplifying that complexity, and presenting coherent, visible career pathways, is a core part of what Tech First is trying to do. 

Aubree added a perspective from her own experience, including from the US, where workplace pilots for incoming fab facilities recruited dental technicians, nail technicians, beauty technicians, and life science lab workers — people with fine motor skills, comfort working under magnification for long periods, and experience in precise, controlled environments. Given clean room readiness training, many thrived, and most reported valuing the stability and predictability the work offered. The point was clear: the semiconductor sector has more lateral entry routes than it typically communicates, and Tech First should be a vehicle for surfacing those. 

The four parts of Tech First 

Cian set out the programme's structure. The first strand focuses on 11–18 year olds — competitions, local activities, and engagement designed to inform subject choices at the critical pre-GCSE and A-level stage. The second is the Tech Grad Programme, which will support up to 1,000 undergraduate and postgraduate scholarships across the frontier industries, each paired with guaranteed practical summer placements with employers. The third is Tech Expert, targeting domestic PhD students and researchers transitioning into frontier industries. The fourth is Tech Local, focused on the demand side — helping SME employers de-risk the process of taking on new entrants from outside the sector, including people moving laterally from other fields. 

Cian was clear that the programme must integrate with what already exists in semiconductors — including the existing undergraduate scholarship scheme and the STEP programme — rather than duplicating it. The goal is a seamless, joined-up offer with as much reach and impact as possible. 

Workplace culture, flexibility, and inclusion 

Aubree broadened the conversation to workplace environments, noting that skills alone are not enough if the conditions for retaining people are not in place. She raised the question of flexibility — acknowledging that fabrication environments require predictable scheduling, but arguing that this doesn't preclude conversations about what working arrangements can look like for a working parent or a neurodiverse individual. She asked what Tech First recommends to employers participating in the programme to ensure they engage with genuine intent, not just to tick a box. 

Cian pointed to the experience of Cyber First employers who had successfully employed neurodiverse individuals and then helped advise others — a model of peer learning he wants to replicate in Tech First. He was candid that the programme won't have all the answers at launch, including on questions like whether aptitude testing inadvertently disadvantages neurodiverse applicants. The commitment, he said, is to keep asking those questions throughout the four years, not to assume they've been resolved. 

Aubree closed this thread with a message she returns to across every audience she works with: what you practice grows stronger. She encouraged everyone in the room to think not just about credentials, but about behaviours — what behaviours Tech First programmes reinforce, and whether those are visible to parents and students. She also put a specific challenge to the room: look at who's being hired in your organisation, and whether they're being assessed on behaviours as well as credentials. 

Mentoring across the career journey 

On the question of mentoring — particularly for early career researchers and PhD students — Cian argued that mentorship needs to run through every strand of the programme, not just one part. Mentors for 11–18 year olds helping them understand what a career could look like; mentors for undergraduates helping them navigate the transition into the workplace; mentors for PhD students helping them understand how to progress in research; and cohort mentors for groups from different communities across the UK. No single person can do all of it, he said, but real-world human connection remains the single most effective lever for helping people enter and progress within the sector. 

Closing messages 

Cian's call to action was direct: engage with Tech First. The programme is being built now, and the sector's input — in designing aptitude tests, taking on placements, and shaping the curriculum — is essential to making it work for semiconductor employers specifically. Membership of the Tech First network is open, and he encouraged everyone in the room to join. 

Aubree left the audience with a series of personal challenges: look at who's being hired in your organisation and whether they're judged on behaviours as well as credentials; ask what you can personally do to close the gender pay gap; consider who you're bringing into the room with you when you come to events like this; invite teachers into your facilities so they can connect the dots for their students; and find ways to reach parents and young people at a community level — not just through formal programmes, but in everyday spaces. The foundations being built now, she said, will determine what the next frontier looks like. 

• Kir Nuthi, Head of AI and Data, techUK (Chair) |  
• Rodolfo Rosini, CEO, Vaire Computing  
•  Raj Gawera, COO, UK Semiconductor Centre 
•  Jeff Kettle, Professor in Electronic Engineering, University of Glasgow / REACT  
• Ramy Shelbaya, Co-Founder & CEO, Quantum Dice  
• Elizabeth Patterson, Senior Policy & Programme Manager, Seagate 

This panel brought together five perspectives from across the semiconductor and emerging technology landscape to address a question that ran through the day: how does a somewhat fragmented sector make a clearer, more compelling case to government — and how do semiconductors get recognised not just as a frontier technology, but as the foundational layer on which the UK's deep tech ambitions rest? 

Sustainability, resilience, and critical materials 

A recurring theme was the link between sustainability and geopolitical resilience. All 18 elements listed by the Critical Minerals Intelligence Centre as critical to the UK economy are used in electronics,  meaning that recovering and securing those materials is as much a supply chain security question as an environmental one, particularly as EU and US legislation increasingly pushes manufacturers to decarbonise. 

The challenge of communicating a fast-moving field to policymakers 

There was sympathy for policymakers navigating a proliferating landscape of competing compute paradigms — quantum, probabilistic, neuromorphic — each with its own advocates and funding bids. The concern raised was that when one approach gains political momentum, it can become self-fulfilling, crowding out alternatives that may be better suited to certain applications. The call was for more honest dialogue that acknowledges uncertainty rather than overselling particular bets, keeping multiple paths viable rather than inadvertently closing them off. 

The breadth and fragmentation of the semiconductor sector itself compounds this challenge — geographically, technologically, and across the supply chain. Bridging the gaps between government, industry, and academia, and helping identify where the UK has competitive strength, is increasingly urgent work. 

Photonics and systems thinking as competitive advantages 

A consensus emerged around photonics as an area of credible, under-recognised UK strength — notably absent from both the AI Opportunities Action Plan and the Compute Roadmap, a gap the panel felt needs addressing. Photonic integrated circuits represent one area where the UK could realistically build an end-to-end domestic ecosystem. Power electronics and compute architectures were cited as further traditional strengths with clear application pathways. 

More broadly, the panel argued that semiconductors should be understood as components of complex systems, not just devices in isolation. The UK's strength in systems thinking — the ability to integrate and optimise at a whole-system level — is a competitive advantage that tends to be underplayed. 

Speed, scale, and the honest assessment 

The panel's most direct challenge to the sector was on ambition. The UK's problem, it was argued by one panellist, is not necessarily a lack of money — it is a lack of speed and strategic commitment. Subsidies and programmes exist, but without a domestic public market to close the investment loop, private capital cannot complete the cycle. The absence of a UK equivalent to the US national labs — each mandated to push computational capability forward — was cited as a structural gap. A recent attempt to showcase British startups at a major government institution attracted no engagement, not from bad intent, but because no mandate or budget existed to act on it. 

The conclusion drawn was unless the UK makes an explicit, measurable commitment to being first or second in specific fields, the activity happening across the ecosystem — however valuable — lacks the scale of ambition needed to compete where it matters. 

Where value gets captured 

One of the sharpest observations of the session concerned where economic value in the AI economy is actually captured. A high-end AI chip sells for tens of thousands of dollars; the raw materials cost a few hundred. The remainder flows to the chip designer. If the UK does not produce the next generation of compute architecture — whether photonics-based or otherwise — the overwhelming majority of economic value from AI processing will effectively function as a tax paid to whoever does. This framing made the case for the UK needing to own a node in the compute value chain, not simply participate in it. 

Closing calls to action 

Each panellist was asked for one concrete action the sector should take forward.  

1.  The first was to identify where the UK can  lead and then focus on removing barriers to private investment — energy costs chief among them — rather than attempting to fund everything from the public purse.  

2. The second was a call for bigger bets: when resources are constrained, outsized outcomes require outsized risk tolerance, not incremental hedging.  

3. The third pushed for a shift toward permissionless innovation — a regulatory and cultural environment where building is the default, not the exception.  

4. The fourth distilled the session's central tension into a single directive: build consensus around focus, pick the areas, get alignment, and act rather than debate. 

5.  The fifth made the case for longer-term, stable funding cycles — some deep tech companies take twelve to thirteen years from startup to revenue, and one- or two-year funding windows are structurally misaligned with those timescales.