AI Regulation Gaps in King's Speech Spark Concerns

The King's Speech, delivered May 22, 2026, laid out the UK government's legislative agenda, and left out any substantial plan for regulating artificial intelligence, an omission advocacy groups say is dangerous given AI's growing role in violence against women and girls (VAWG).

What Happened

The speech's silence on AI regulation stands out against the backdrop of AI-driven surveillance and facial recognition tools already in use, technologies advocacy groups have criticized for privacy risks and for biases that can worsen discriminatory outcomes like racial profiling.

The omission is notable specifically because of the VAWG angle: unregulated AI systems can be used to track, predict, or exploit user behavior in ways that disproportionately put women and girls at risk, and the speech offered no framework to address that.

How This Affects Everyday People

Without regulation, the immediate risk runs in two directions. Workers in retail and manufacturing face job insecurity as AI-driven systems take on more tasks without any legal requirement for companies to plan for that transition. Parents, meanwhile, have no new statutory backstop for how AI-powered platforms track and predict their children's behavior online.

The VAWG dimension is the sharpest concern: AI systems that aren't regulated can encode and amplify existing biases into hiring, law enforcement, and healthcare decisions, exactly the kind of unchecked deployment critics say the King's Speech should have addressed and didn't.

The Bigger Picture

The UK's gap is thrown into relief by the EU's own AI Act, which does set a comprehensive regulatory framework, meaning the UK is now behind a major neighboring jurisdiction on this specific issue. That gap is feeding a growing backlash against AI, as advocacy groups press for the UK to close it.

What You Can Do

• Track UK vs. EU AI policy: see how the EU AI Act compares to what (little) the UK has proposed.
• Support VAWG and AI-safety advocacy groups pushing for the UK to close this specific regulatory gap.
• Contact your MP to raise the absence of AI regulation in the King's Speech directly.
• Choose platforms with transparent AI practices where you have the option, especially for children's accounts.

The Bottom Line

Leaving AI regulation out of the King's Speech isn't a neutral gap. It leaves the VAWG risks of unregulated AI systems unaddressed while the EU already has a framework in place. Closing that gap now depends on public pressure, since the government's own agenda-setting speech didn't do it.

How the UK's Gap Compares

Our own tracker of AI-related legislation and enforcement actions gives some sense of scale for what the King's Speech left out. We've logged 106 distinct AI policy actions since 2008, 62 of them binding legislation, the rest company policies, worker-protection measures, pledges, and contracts, and the overwhelming majority are American state and city laws: New York City's bias-audit requirement for hiring algorithms (Local Law 144), California's data-center water-and-emissions disclosure law, and dozens more. The UK's King's Speech adding zero new binding measures to that list, in a year when other jurisdictions added several, is what makes the omission notable rather than routine.

Separately, our AI lawsuits tracker has logged 47 active cases, 28 of them copyright claims and 5 specifically privacy claims, the exact category most relevant to the VAWG risk this story raises, since AI systems that track or predict user behavior without regulation are also the systems most likely to end up as privacy defendants. The absence of a UK regulatory framework doesn't stop the litigation risk; it just means the courts, rather than Parliament, end up setting the rules first.

The EU AI Act itself takes a tiered, risk-based approach: it bans a small set of "unacceptable risk" uses outright, like social scoring, imposes strict transparency and oversight obligations on "high-risk" systems (including many hiring and law-enforcement tools), and leaves low-risk applications largely unregulated. That structure is what UK advocacy groups are pointing to when they argue the King's Speech should have proposed something similar rather than nothing. A comparable framework doesn't require banning AI, just tiering the obligations to the actual risk a given use case carries.

The practical gap for UK readers right now is enforcement, not just legislation: without a binding framework, complaints about biased or exploitative AI systems have no statutory body to escalate to inside the country, which is why advocacy groups keep pointing across the Channel rather than to Westminster. Until that changes, the EU AI Act functions as the de facto reference point for what "good" AI regulation looks like, even for UK citizens it doesn't legally cover.

What Existing UK Law Already Covers

The King's Speech didn't leave the UK with zero legal tools against harmful AI, even without a dedicated AI Act. The UK's data protection regulator, the Information Commissioner's Office (ICO), already has authority under UK data protection law to investigate automated decision-making that carries real consequences for people, including profiling.

The Equality Act 2010 covers discriminatory outcomes from an algorithm much the way it covers discrimination from a human decision-maker, at least in principle. The law wasn't written with AI in mind, but the general protection still applies. The UK has plenty of legal tools that touch AI indirectly, just none built for it specifically.

An AI-specific framework would name the risks in advance. That beats leaving regulators to stretch general-purpose laws onto a new kind of harm only after something has already gone wrong.

Why a Law on the Books Isn't the Same as Protection in Practice

Having a framework like the EU AI Act doesn't automatically mean every risk it names gets caught in time. The Act phases in its obligations over several years, and the highest-risk category rules carry the longest lead times before they're fully enforceable.

That design gives companies time to comply. It also means some of the Act's strongest protections aren't fully active yet, even inside the EU.

The UK's absence of any framework is a different problem than the EU's slow rollout of one. Both point at the same gap: the time between a risk becoming visible and a regulator actually being able to act on it.

What Advocacy Groups Say Should Replace the Gap

The asks that keep surfacing across VAWG and AI-safety advocacy groups follow a similar shape:

• Mandatory bias audits for AI systems used in policing, hiring, or benefits decisions.
• A statutory body with enforcement power specifically over AI harms, rather than one stretched to cover AI on top of an existing mandate.
• A reporting requirement so incidents of AI-enabled harm get tracked centrally instead of surfacing case by case in the press.

None of those proposals require banning AI outright, which is part of why advocates point to the EU's tiered approach as evidence the ask is workable, not just aspirational.

Questions Readers Are Asking About the UK's AI Policy Gap

Is the UK planning any AI legislation at all? The King's Speech didn't include one, which is the specific gap this story covers. That doesn't rule out separate proposals moving through Parliament on a different timeline.

Does data protection law already cover AI systems in the UK? UK data protection law applies to AI systems that process personal data, which covers a meaningful share of AI harms. It wasn't written with AI-specific risks like algorithmic bias or VAWG-linked exploitation in mind, which is exactly the gap advocacy groups are pointing to.

Further Reading

This story is based on original reporting via Google News: "King's Speech: VAWG and AI regulation largely missing."