The United Kingdom is at an important moment in AI policymaking.
During ACT’s Global App Economy Conference: UK in April, one question kept coming up across conversations with policymakers, regulators, and industry voices: how does the UK turn AI ambition into real deployment and growth?
ACT hosted a reception in the House of Commons, bringing policymakers and startups together to discuss innovation, AI, and growth.
The answer starts with policy design. The UK has so far taken a pro-innovation approach rooted in risk-based oversight, sector-specific expertise, and existing law. That approach matters. Rather than drafting a broad horizontal AI law like the EU AI Act, the UK is favouring a model where existing regulators apply their expertise to specific sectors and use cases. This approach should be preserved.
AI policy should stay connected to real deployment. Small businesses are already using AI to move from idea to product faster, improve cybersecurity, develop creative tools, and build services in sectors such as music, space, and government. These examples show AI as a practical tool for solving defined problems, improving workflows, and reaching new users.
Policy should also account for concerns around workforce transition and cybersecurity. These issues deserve focused responses through education, reskilling, and practical safeguards, rather than broad rules that treat every AI use case the same way.
A risk-based framework gives policymakers room to address real harms while preserving space for beneficial deployment. Blanket premarket audits, rigid documentation duties, or broad obligations detached from use cases would fall hardest on smaller companies. Large firms have legal teams and compliance departments. Most startups and scaleups do not. For them, uncertainty is a cost, and excessive process delays growth.
Copyright is one area where uncertainty is already shaping business decisions. The issue sits at the centre of the UK’s AI strategy because developers, rightsholders, investors, and creative businesses all need clearer rules. The UK’s decision to preserve the status quo and avoid a new commercial text-and-data-mining exception gave publishers and rightsholders a near-term win. But the current position still leaves developers and investors without a predictable path for commercial AI training.
ACT members meeting with policymakers to discuss AI, innovation, and small business growth.
That balance matters. A framework with too little clarity chills investment. A framework built around costly licensing mandates favours the largest firms. Smaller developers need rules they can understand and apply without bespoke legal advice at every step.
Funding and scale present a second test. The UK has strong startup activity and a clear opportunity to help more companies scale. But funding remains fragmented, and investment remains uneven across regions. A stronger focus on regional startups would help promising firms with clear market value access the capital they need to grow.
These concerns matter because AI companies need compute, talent, infrastructure, partnerships, and customers. The UK’s new sovereign AI fund reflects serious attention to this challenge. The government has launched a $675 million venture fund to invest in domestic AI startups and reduce dependence on foreign-made technology. That investment sends a useful signal. Public capital should reach across the AI ecosystem. The UK needs to provide more support for companies building on top of frontier models and developing application-layer tools.
Public procurement is another important lever. Government adoption of AI should be part of this strategy. AI systems can improve public service delivery, make services more efficient, and expand access for British citizens. Procurement policy should make room for smaller firms with practical products, rather than only large vendors with mature compliance teams.
Standards and interoperability matter for the same reason. Standardised interoperability is a foundation for innovation, safety, and competition. Small businesses and startups need certainty that they can use these standards as they build and plan, specifically that promises made by standard-essential patent (SEP) holders to provide licences to all on fair, reasonable, and non-discriminatory (FRAND) terms will be kept.
The UK has an opportunity to take a practical path. It has avoided broad horizontal AI regulation. It is investing in compute, growth zones, public sector adoption, and domestic AI capacity. It is considering tools such as the AI Growth Lab to help products reach the market. These choices point in the right direction.
ACT staff and members at the close of a successful 2026 UK Global App Economy Conference.
The next step is increasing certainty. The UK should focus on clearer copyright rules, risk-based oversight, startup-friendly funding and procurement, holding SEP holders to the FRAND promises they make, and avoiding broad regulatory models that add compliance costs without improving outcomes.
The UK can support responsible AI governance and innovation through a framework that helps both scale together.