The United Kingdom faces a critical juncture on AI ethics and regulation, with no settled strategy in place even as the European Union has already laid down the world’s first comprehensive legal framework governing artificial intelligence systems.
AI Ethics Failures Have Already Reached UK Streets
The urgency is not theoretical. In 2020, the UK government deployed an AI algorithm to replace cancelled A-Level exams, only for the system to systematically under-predict the grades of students from underprivileged backgrounds. Students took to the streets carrying banners reading “Your algorithm doesn’t know me,” according to reporting on the episode. Separately, the Home Office was successfully challenged in court over its use of an algorithm to sift visa applications — a system critics said amounted to “speedy boarding for white people” by entrenching racially biased decision-making. South Wales Police also faced a successful human rights challenge over its use of facial recognition technology in public spaces.
These incidents have intensified calls for clear statutory guardrails. They also illustrate a pattern that AI literacy advocates and legal experts say will repeat itself without structural oversight — a concern that goes to the heart of ongoing debates about how much autonomous decision-making AI should be trusted with.
The EU’s AI Regulation Sets a Global Benchmark
In March 2021, the EU became the first political system in the world to define an all-encompassing legal framework on artificial intelligence. The proposed regulation is built around a risk-tiered structure. Systems deemed to pose an “unacceptable risk” — including those that manipulate behaviour likely to cause psychological or physical harm, exploit children, or deploy Chinese-style social scoring — would be banned outright, according to the EU’s published proposals.
High-risk systems, covering AI used in critical infrastructure, medical devices, recruitment, credit scoring, and biometric identification, would face strict requirements around data governance, record-keeping, transparency, and human oversight. Lower-risk systems such as chatbots would face lighter obligations, including a requirement to disclose to users that they are interacting with a machine.
The European Commission stated that the framework aims to “guarantee the safety and fundamental rights of people and businesses, while strengthening AI uptake, investment and innovation across the EU.” Policy observers at the European Parliament have described it as a foundational moment in technology governance globally.
However, critics argue the regulation does not go far enough on AI ethics. Dr Claudine Tinsman, policy adviser at the non-profit We and AI, points out that while Article 5 of the draft regulation restricts the use of remote biometric identification in publicly accessible spaces, it applies only to real-time systems used specifically for law enforcement. Non-live facial recognition — which can still identify individuals after the fact — is not prohibited. The regulation also does not ban real-time identification of people suspected of irregular entry or residence, a provision that critics say could disproportionately expose migrants and refugees to surveillance.
More than 60 non-governmental organizations called on the European Commission to ban AI systems that automatically identify gender and sexual orientation. That prohibition was not included in the draft. Nor was a ban on emotion-sensing AI — systems that purport to detect psychological states through micro-expressions and physiological signals. Such tools have reportedly been deployed at events including the FIFA World Cup and the Olympic Games, as well as at US and EU border crossings, despite what researchers describe as an absence of credible evidence for their accuracy. Under the EU framework, emotion-sensing AI is classified as high-risk but remains permitted.
“Facial recognition systems, in particular, are biased against minorities, so leaving the door open for their use on a large scale could disproportionately affect minorities,” Tinsman said.
The UK Must Now Choose Its Own AI Ethics Path
Outside the EU since Brexit, the UK is not bound by the bloc’s regulation and must determine its own approach. In January 2021, the AI Council delivered a Roadmap of recommendations to the government’s Office for AI, targeting an increase in UK GDP by 2030 while also promising benefits for the environment and people from “all walks of life.” How the government translates those recommendations into law — or whether it does so at all — remains unresolved.
Minesh Tanna, managing associate and AI lead at law firm Simmons and Simmons, told the author that full adoption of the EU framework is unlikely given its strictness and the UK’s need to attract inward investment. A middle path is more probable: voluntary conformity assessments for UK AI companies combined with adoption of the EU’s outright prohibitions. “The UK might also want to use its flexibility to regulate in a more tailored, nuanced way — an advantage of Brexit — while at the same time, probably having a less onerous regulatory regime than the EU, so as to encourage investment,” Tanna said. Without some form of statutory framework, Tanna warned, the UK risks being seen as a destination for “ethics dumping” — a place where companies deploy AI systems that would be prohibited elsewhere.
Charles Radclyffe, AI ethics and ESG specialist and partner at EthicsGrade, cautioned that the UK also needs to ensure sufficient capital investment stays within its AI ecosystem rather than being drawn toward the US or China. “The UK is a little like Pluto. We’re still in orbit of the sun [the EU] but on the face of it we have lost all of the benefits of planetary status,” Radclyffe said. “What we need to do is create the conditions and highlight sufficiently the uniqueness of the UK’s AI ecosystem.”
Tania Duarte, who founded the AI literacy non-profit We and AI in 2020, argues that the scrutiny of AI ethics cannot be left to companies or governments alone. Greater public awareness, she contends, is essential for people to identify violations of existing equality and privacy laws and to participate meaningfully in debates about the trade-offs built into algorithmic systems. The EU has funded programmes to increase AI ethics literacy among the general population; the UK has no equivalent at present.
That gap matters. As concerns about AI’s societal impact continue to grow, particularly in public institutions, the absence of both a statutory framework and broad public understanding leaves citizens poorly equipped to hold either governments or corporations to account. The decisions the UK makes — or delays — in the coming months will shape how AI systems affect millions of people long before any corrective legislation could catch up. Researchers tracking data governance and security risks warn that governance vacuums consistently invite misuse, and AI is no exception.
The EU, for all its critics, has at least planted a legal stake in the ground. Whether the UK follows, adapts, or ignores that precedent is now among the most consequential technology-policy questions it faces, according to the UK government’s own advisory bodies.
Key Takeaways
- The UK has no settled AI regulation strategy despite documented failures of government-deployed algorithms that harmed minority and low-income communities.
- The EU’s March 2021 AI regulation — the world’s first comprehensive framework — bans the highest-risk systems but contains significant loopholes on facial recognition and emotion-sensing AI, according to critics including We and AI policy adviser Dr Claudine Tinsman.
- Legal experts expect the UK to pursue a lighter-touch regime than the EU’s, combining voluntary assessments with adoption of outright prohibitions, in order to attract investment while avoiding the label of an “ethics dumping” destination.
- AI ethics advocates argue that public literacy — not just legislation — is essential, as it enables citizens to identify breaches of existing equality and privacy law and participate in debates about algorithmic trade-offs.
- The UK’s post-Brexit position means it must define its own AI ethics framework independently, a decision that will have lasting consequences for civil rights, innovation, and international competitiveness.
