Commons Sense

House of Commons · Ministers' Questions

AI: Regulation

Wednesday 17 April 2024 · 8 questions

Summarised by AI from the official record, so it can contain mistakes.

Read the full transcript on Parliament.uk

At a glance

Answered 1 Partly answered 3 Not answered 0

Key points

  • The White Paper on AI regulation outlines five cross-sectoral principles but lacks concrete actions and timelines for implementation.
  • The UK is leading on AI safety through initiatives like the AI Safety Summit and the world's first AI Safety Institute.
  • The Government is committed to engaging internationally on AI regulation, including a memorandum of understanding with the US on AI institutes.

Topics (select to filter)

Questions & Answers

Q1 Henry Smith Con
Science & Technology
Context

The UK government is under pressure to regulate artificial intelligence as concerns about its impact grow.

Question

What steps her Department is taking to regulate artificial intelligence. The White Paper on AI regulation outlines five cross-sectoral principles but lacks concrete actions and timelines for implementation.

Answer from Michelle Donelan

In our White Paper on AI regulation we set out an ambitious pro-regulation, pro-innovation framework outlining five cross-sectoral principles. We have established a central risk function and provided new funding and guidance for regulators.

Partly answered

Not addressed: Specific timelines or concrete actions not addressed

Was this summary accurate? Tell us

How accurate was this summary? (1 = poor, 5 = spot on)

Get updates like this by email
Q2 Chris Elmore Lab Bridgend
Science & Technology
Context

The Competition and Markets Authority has warned about the potential harms from unregulated AI, including biases affecting certain demographics.

Question

Why have the Government dragged their feet on safeguards for advanced AI models? The Prime Minister previously stated that rushing to regulate is not the answer. Are they waiting for a Labour government?

Answer from Michelle Donelan

We are leading on AI safety with initiatives like the AI Safety Summit and establishing the world's first AI Safety Institute, testing both pre and post deployment. We will not rush to legislate but want to understand risks better.

Not addressed: Specific timeframe for safeguards avoided

How: Changing Subject To International Initiatives

Was this summary accurate? Tell us

How accurate was this summary? (1 = poor, 5 = spot on)

Get updates like this by email
Q3 Justin Madders Lab Ellesmere Port and Bromborough
Employment Science & Technology
Context

The TUC will launch a Bill on AI regulation and employment rights, emphasizing transparency in the workplace.

Question

What measures will the Government take to ensure that AI does not weaken workers' rights? The TUC's Bill focuses on transparency, observability, and explainability as key elements of fair workplaces.

Answer from Michelle Donelan

We are focused on garnishing opportunities from AI for jobs in sectors like teaching and medicine. We acknowledge that technology changes labour market needs, hence our skills system revolution including a lifelong learning entitlement next year.

Not addressed: Worker rights protections not addressed

How: Shifting Focus To Job Opportunities

Was this summary accurate? Tell us

How accurate was this summary? (1 = poor, 5 = spot on)

Get updates like this by email
Q4 Nia Griffith Lab Llanelli
Science & Technology
Context

The Federation of Small Businesses calls for an urgent regulatory framework, citing potential delays in AI regulation.

Question

Can the Secretary of State provide a timeframe for AI regulation? The FSB and Dr Rogoyski highlight urgency due to risks of delay.

Answer from Michelle Donelan

We are committed to unlocking opportunities while mitigating risks, engaging internationally with the US on AI institutes through a memorandum of understanding. We will legislate carefully and globally, as others do.

Not addressed: Specific timeframe not provided

How: Discussing International Cooperation

Was this summary accurate? Tell us

How accurate was this summary? (1 = poor, 5 = spot on)

Get updates like this by email
Q5 Henry Smith Con
Foreign Affairs Science & Technology
Context

The UN plans to internationally regulate artificial intelligence, raising questions about UK sovereign security.

Question

What assessments have the Government made of the United Nations' international regulation plans for AI? Implications for UK sovereignty must be considered.

Answer from Michelle Donelan

The UK is committed to unlocking AI opportunities while mitigating risks, both domestically and internationally. We actively engage in the UN and shape UNESCO recommendations on AI ethics.

Not addressed: Specific assessments not provided

How: Discussing International Engagement

Was this summary accurate? Tell us

How accurate was this summary? (1 = poor, 5 = spot on)

Get updates like this by email
Q6 Greg Clark Con
Science & Technology
Context

Leading AI developers are expected to release new models, and the AI Safety Institute has access for pre-deployment testing.

Question

Has our AI Safety Institute had access to upcoming AI models? Developers have made changes as requested. What specific actions have been taken?

Answer from Michelle Donelan

Our institute is the first in the world to test pre and post-deployment, including access agreed at Bletchley Park. Specific model details are commercially sensitive.

Partly answered

Not addressed: Specific model details not disclosed due to sensitivity

How: Citing Commercial Sensitivity

Was this summary accurate? Tell us

How accurate was this summary? (1 = poor, 5 = spot on)

Get updates like this by email
Q7 Oliver Heald Con
NHS Science & Technology
Context

AI has potential benefits for public health, such as diagnosing bowel cancer.

Question

Does my right hon. Friend agree that promoting best practices and accelerating healthcare missions with AI is urgent? Simply slowing things down would be counterproductive.

Answer from Michelle Donelan

Absolutely, AI can revolutionize healthcare. We announced a £100 million pot to accelerate existing healthcare missions and are working closely with the Department of Health and Social Care.

Answered
Was this summary accurate? Tell us

How accurate was this summary? (1 = poor, 5 = spot on)

Get updates like this by email
Q8 Matt Rodda Lab Reading Central
NHS Democracy & Elections Science & Technology
Context

The Government have been warned about the risks associated with AI technology such as deepfakes. There is criticism that Ministers have delayed action on these issues.

Question

AI is an incredible new technology and can help the NHS to save lives, but there are also risks, such as the danger of deepfakes. The Government have been warned about those risks, yet time and again Ministers have dithered and delayed. Have the Government run out of ideas, or are they just scared of their own Back Benchers?

Answer from Michelle Donelan

We have the defending democracy taskforce, which is dedicated to this subject and led by the Security Minister. We work with social media companies and our international counterparts. The Deputy Prime Minister leads the way with his AI compact. There is no easy answer, but we are working conciliatorily and quickly to address all opportunities and risks.

Partly answered

Not addressed: The specific ask about whether the Government has run out of ideas or fears acting was not directly addressed.

How: Hedging Language, Broad Statements

Was this summary accurate? Tell us

How accurate was this summary? (1 = poor, 5 = spot on)

Get updates like this by email