Creative and marketing agencies could be required to disclose when they use AI, explain what it is doing and accept greater responsibility for preventing harm under proposals for a new UK AI Bill.
Agencies and other creative businesses could face mandatory disclosure, due diligence and accountability requirements under proposals published by Parliament’s Joint Committee on Human Rights.
The committee’s new report, Human Rights and the Regulation of AI, calls for a dedicated AI Bill and an independent regulator with powers to investigate businesses, impose sanctions and prevent high-risk systems from being released.
The recommendations have not been adopted by the government and do not represent existing legal duties. The government now has two months to respond to the report.
Lord David Alton, the former Liberal MP for Liverpool Mossley Hill who led the inquiry during his final period as chair of the committee, described the unanimously agreed report as both timely and important.
Writing on X, Alton said: “My last act as Chair of @UKParliament JCHR was to oversee the Inquiry and Report into Human Rights and the Regulation of Artificial Intelligence. This Report could not be more timely or important.”
Alton, who has now stepped down as chair and been succeeded by Alex Sobel MP, said the committee had received more than 70 pieces of written evidence and held ten oral evidence sessions.
Those contributing included academics, researchers, campaigners, legal and technology specialists, entrepreneurs, major technology companies, regulators and the AI minister.
READ MORE: Prolific North Live 2026: a first look at the brands shaping the agenda
Alton added: “In order to clearly set out the case for change, this report starts by explaining what artificial intelligence is and why we believe it raises vital human rights issues. We then discuss the existing legal and regulatory protections, and explain why we consider they are seriously lacking. Finally, we set out our proposals for comprehensive, tailored law and regulation to address these risks and protect human rights.”
The peer, who has sat as an independent crossbencher since 1997 and is a visiting professor at Liverpool Hope University, said the report had arrived amid escalating warnings from some of the world’s most prominent AI leaders.
He wrote: “If you have any doubt about the dangers (as well as opportunities) posed by AI, read why the co-founder of Anthropic has called for a pause in the development of AI – naming the threat it poses to the world.”
AI leaders back calls to slow development
Alton’s comments follow a rare show of agreement involving senior figures at rival AI companies.
As reported by The Guardian, Anthropic founder and chief executive Dario Amodei has called on the industry to “slow the pace” of development and introduce independent monitoring.
Amodei warned that “building [AI] too fast is reckless” and raised the prospect of increasingly capable AI agents causing widespread cyber damage.
His assessment has been disputed by some AI experts, who questioned the plausibility of the scenario and said it should be treated with scepticism.
However, OpenAI chief executive Sam Altman backed the broad direction of Amodei’s proposal.
Altman said: “I agree with Dario that we need to pace the frontier.”
Elon Musk responded separately: “Dario is right.”
Google DeepMind co-founder and chair Demis Hassabis also said: “The details need working through, but the direction is correct for meeting this critical moment.”
Alton wrote on X: “The JCHR Report is timely and its reciommendations require an urgenbt response.”
New transparency requirements
Under the committee’s proposed regime, organisations using AI systems with a significant impact on individuals, groups or communities would have to state when the technology was being used and provide a “full and comprehensible explanation” of its purpose.
Businesses could also be required to reveal the source of data used by an AI system.
For agencies, the proposals could affect areas including creative production, audience targeting, personalisation, recruitment, casting, customer communications and the development of AI-powered products for clients.
The committee said responsibility should extend throughout the AI supply chain, covering those designing, developing and deploying systems.
Its report argues that large technology companies currently have too much freedom to pass liability to businesses using their products, despite those customers frequently having less technical knowledge and limited access to information about how the underlying models work.
The proposed AI Bill would instead place proportionate obligations on the organisations best positioned to identify and prevent particular risks.
That could mean agencies being accountable not only for how they use third-party AI platforms, but also for AI applications, automated services and customised systems they build or supply to clients.
High-risk providers and users could be required to conduct due diligence including risk assessments, testing, data governance and measures to prevent potential human-rights harms.
The report also says the presence of a “human in the loop” should not automatically be treated as meaningful oversight. The individual reviewing an automated decision would need to be sufficiently informed and independent to reach an objective conclusion rather than merely approving the system’s recommendation.
Sobel, the committee’s new chair, said: “AI is heralded as an unprecedented era of technological development with the potential to transform our lives for better or for worse. It is moving with such speed and complexity that its impact is hard to accurately predict. What is clear is that at present we are unprepared to deal with its consequences however potentially dire they may be.
“Nowhere in the world, including the UK, has a current legislative and regulatory approach to AI that is fit for purpose. New legislation is needed to establish a comprehensive set of protections that deal with the entire AI supply chain and its lifecycle. A single AI regulator should be established to set policy, monitor performance and with the teeth to ensure enforcement.”
Creative rights require further investigation
The report raises particular concerns for creative businesses over copyright, performers’ rights, synthetic content and biased decision-making.
Evidence submitted to the inquiry warned that copyrighted and professional creative work was being incorporated into AI training datasets, while artists could find it extremely difficult to establish whether their material had been used.
The committee did not make detailed recommendations on copyright because intellectual property was not a central focus of its inquiry. It concluded, however: “The issues which witnesses have raised clearly merit further investigation, and could require specific measures going beyond the scope of this inquiry.”
It added that it was important to protect the UK’s creative industries and safeguard performers’ human rights and livelihoods.
The report cites evidence from Equity concerning a performer who had agreed to record a voiceover for non-commercial educational material intended to help visually impaired readers.
According to the union, the performer subsequently discovered that a cloned version of her voice had been “made available to others via a text-to-speech tool […] without [her] consent, control or pay”.
The report also identifies the use of potentially biased AI systems in creative-industry casting as a discrimination risk.
Other proposed restrictions could affect businesses working in advertising technology, audience analysis and workplace monitoring. The committee recommends consulting on prohibitions covering areas such as subliminal techniques, emotional inference and inappropriate profiling or use of biometric data.
An independent AI oversight body would be able to publish mandatory codes of practice, investigate suspected breaches, sanction developers and deployers, and order remedies for people affected by AI-related harm.
It could also prevent an AI model or system from being launched or require its withdrawal where unacceptable risks were identified.
The committee additionally wants the AI Security Institute placed on a statutory footing and given the power to review new and revised powerful AI models before their release.
Developers would be required to provide technical specifications covering areas including model properties, training data, intended uses, safety testing and risk-control measures.
However, the committee said the eventual regime should be proportionate. Lower-risk applications would face less demanding requirements, while legislation should avoid placing unjustified burdens on smaller businesses.
It argued that properly targeted regulation would not prevent agencies and technology companies from innovating, but would create clearer responsibilities and give individuals a way to challenge damaging automated decisions.
