Last reviewed
Jun 7, 2026
Sources
20 citations
Review status
Source-backed
Revision
v1 · 2,072 words
Add missing citations, update stale details, or suggest a clearer explanation.
The UK Regulating for Growth Bill is a legislative proposal announced in the United Kingdom's King's Speech on 13 May 2026, setting out the Labour government's plan to make economic growth a central duty of regulators and to create cross-economy "sandbox" powers that let firms trial new products, including artificial intelligence tools, under temporarily modified rules with regulator oversight. The bill is the legislative vehicle expected to carry the AI Growth Lab, a programme of issue-specific regulatory sandboxes put out to consultation by the Department for Science, Innovation and Technology (DSIT) in October 2025. As of June 2026 it has been announced but not introduced, debated, or enacted; the detail summarised here reflects stated intentions and is subject to change as a bill is drafted. The proposal confirms that the UK is continuing its sector-led, pro-innovation course on AI regulation rather than adopting a single horizontal AI statute on the model of the EU AI Act, and that a long-discussed dedicated AI bill remains deferred.
The UK's distinctive stance on AI dates to the March 2023 white paper "AI regulation: a pro-innovation approach," which deliberately avoided a single AI law. Instead, it asked existing sector regulators, such as the Information Commissioner's Office (ICO), the Financial Conduct Authority (FCA), the Competition and Markets Authority (CMA), and Ofcom, to interpret and apply five non-statutory cross-sector principles within their own remits: safety, security and robustness; appropriate transparency and explainability; fairness; accountability and governance; and contestability and redress. This sectoral, principles-based model was framed as more agile and less burdensome than prescriptive legislation, and it remains the foundation of UK AI policy.
After the July 2024 general election, the Labour government retained the pro-innovation framing but shifted emphasis toward growth. The AI Opportunities Action Plan, launched on 13 January 2025 and endorsing all 50 recommendations of a review led by Matt Clifford, reframed the role of sector regulators: rather than simply guarding against harm, they were now expected to actively promote AI adoption to drive economic growth. A January 2026 progress report, "One Year On," reported that 38 of the 50 commitments had been met. During this period the government also rebranded the AI Safety Institute as the AI Security Institute on 14 February 2025, with Technology Secretary Peter Kyle announcing at the Munich Security Conference a "renewed focus" on national security threats such as AI-enabled cyber attacks, chemical and biological weapons, fraud, and child sexual abuse material. The institute, a directorate of DSIT, kept the acronym AISI and continues to run pre-deployment evaluations of frontier models.
A standalone, comprehensive AI bill was widely expected during this period and was at one stage anticipated for spring 2025. It was repeatedly delayed, in part to fold in unresolved questions about AI and copyright, and did not appear in the 2026 King's Speech. Ministers including Tech Secretary Liz Kendall and AI Minister Kanishka Narayan had signalled in advance that no dedicated AI bill would feature in the session.
The State Opening of Parliament and the King's Speech took place on 13 May 2026, with King Charles III reading the speech from the throne in the House of Lords chamber. The speech set out roughly 37 bills the government hoped to pass in the session. Among them was the Regulating for Growth Bill, framed around the goal of reducing "the burden of unnecessary regulation through innovation."
Crucially, AI was threaded through this growth-and-deregulation bill rather than addressed by separate primary legislation. According to legal and policy analysts who reviewed the accompanying background notes, the Regulating for Growth Bill is intended to support pilot schemes in priority areas of innovation, with sectors such as defence and AI named explicitly. Bird & Bird described it as the first concrete UK statutory framework touching AI oversight since the 2023 white paper, following roughly three years of pro-innovation guidance without primary legislation. Commentators were also clear that no dedicated AI bill was announced, consistent with prior ministerial signalling.
The bill has two broad strands. First, a strengthened statutory growth duty for regulators, described as elevating consideration of growth in regulatory decision-making without undermining regulators' core objectives such as safety or the environment, supported by a new ministerial power to issue "strategic steers" defining what growth means in different regulatory contexts, plus new reporting requirements to measure delivery against the duty. Second, cross-economy sandbox powers allowing businesses to test innovative products and technologies, including AI, medicines, and autonomous and defence technology, in real-world settings under controlled conditions, with successful trials capable of being embedded permanently into law.
The Regulating for Growth Bill is widely understood as the legislative mechanism to underpin the AI Growth Lab, which DSIT had already opened for consultation. The Lab is described on GOV.UK as "a pioneering cross-economy sandbox that would carefully supervise the deployment of AI-enabled products and services that current regulation hinders." It was unveiled by Liz Kendall, then Technology Secretary, at the Times Tech Summit in London on 21 October 2025, the same day DSIT issued a call for evidence. The call for evidence closed at 11:59pm on 7 January 2026, having been listed as closed on the GOV.UK page from 18 December 2025.
Under the proposals, the Lab would grant time-limited regulatory modifications to eligible firms and products, described as "sandbox pilots," and would test these in live market environments under regulator supervision. Successful pilots could then be converted into permanent reforms through updated guidance, codes of practice, or statutory amendments; DSIT is also considering powers to make such permanent modifications by secondary legislation, subject to parliamentary scrutiny. The consultation set out two possible governance models: a centrally operated Lab run by government with an oversight committee of sectoral regulators, or a regulator-led model in which individual regulators or consortia run sandboxes within their remits.
Initial priority sectors map onto the UK's modern industrial strategy, namely advanced manufacturing, digital and technologies, financial services, life sciences, and professional and business services, with specific early use cases including the planning system, NHS diagnostic imaging and radiology, and autonomous robotics and delivery devices. Importantly, the government set "red lines" that could not be modified in Lab testing: consumer protections, safety provisions, fundamental rights, workers' protections, and intellectual property rights.
| Element | Detail (as announced or proposed) |
|---|---|
| Bill name | Regulating for Growth Bill |
| Announced | King's Speech, 13 May 2026 |
| Status | Announced, not yet introduced or enacted (as of June 2026) |
| Lead department | Department for Science, Innovation and Technology (DSIT) |
| Strand 1 | Strengthened statutory growth duty for regulators; ministerial "strategic steers"; new reporting requirements |
| Strand 2 | Cross-economy sandbox powers for AI, medicines, defence and autonomous tech |
| AI Growth Lab call for evidence | Opened 21 October 2025; closed 7 January 2026 |
| Governance options | Centrally operated Lab, or regulator-led sandboxes |
| Priority sectors | Advanced manufacturing, digital, financial services, life sciences, professional services |
| Red lines | Consumer protections, safety, fundamental rights, workers' protections, IP rights |
| Permanent reform route | Conversion of pilots via guidance, codes, or secondary legislation |
The Regulating for Growth Bill underscores how far the UK has diverged from the European Union. The EU AI Act is a horizontal, risk-based law that classifies AI systems into tiers, unacceptable, high, limited, and minimal risk, bans certain practices outright, such as social scoring and some biometric uses, and imposes prescriptive obligations on high-risk systems, overseen by the European AI Office alongside national authorities. The UK has no statutory risk classification, no central AI authority of that kind, and no list of prohibited AI practices in law; oversight instead rests with existing sector regulators coordinated through the Digital Regulation Cooperation Forum, while AISI evaluates frontier models.
The contrast sharpened in May 2026. On 7 May 2026 the European Parliament and Council reached a provisional agreement on a "Digital Omnibus" simplification package that slims down and delays parts of the EU AI Act, pushing certain high-risk obligations to dates in December 2027 and August 2028 depending on the product type. The UK's near-simultaneous choice, six days later in the King's Speech, to legislate for growth and sandboxes rather than for a comprehensive AI regime illustrates two regulatory philosophies converging on the political theme of easing compliance burdens by very different routes. UK firms that place AI on the EU market remain in scope of the EU AI Act regardless of the domestic approach.
As of June 2026 the Regulating for Growth Bill is an announced proposal, not enacted law. It must still be drafted, introduced, and passed through both Houses of Parliament before any sandbox powers or growth duties take legal effect, and the AI Growth Lab's final design depends on the government's response to its call for evidence and on the powers the eventual bill provides. How "strategic steers" interact with regulators' statutory independence will be tested during legislative scrutiny.
Reaction has been mixed. Industry and legal commentators broadly welcomed the focus on agility, with analysts at Pinsent Masons, Lewis Silkin, and Hogan Lovells treating sandboxes as a credible route to faster machine learning and AI deployment, and bodies such as techUK arguing that AI Growth Labs could accelerate innovation. Officials justified the reforms by pointing to a perceived lack of agility in UK regulation compared with the United States, China, Singapore, and Canada. Critics and some legal analysts cautioned about regulatory uncertainty and "stagnation," warning that repeatedly deferring a dedicated AI law leaves frontier developers without a single statutory framework, and questioned whether a strengthened growth duty could pressure regulators to under-weight other objectives. Civil society and academic responses, including from the Responsible AI UK programme, pressed on safeguards and accountability for sandbox modifications.
For now, the Regulating for Growth Bill stands as the clearest statement to date of the UK's intention to use regulatory sandboxes and a growth-oriented regulatory culture, rather than a comprehensive AI statute, as its primary tools for governing AI and emerging technologies, while questions of AI safety and frontier-model oversight remain handled separately through DSIT and the AI Security Institute. Whether the approach delivers growth without weakening protections, and whether a dedicated bill for large language models and other frontier systems eventually follows, will be a central question for UK technology policy through the 2026 to 2027 session.