Artificial Intelligence 2023-07-19
2023-07-19
TAGS
Response quality
Questions & Answers
Q1
Direct Answer
▸
Context
The question arises from the need to regulate AI given its rapid development and impact on various sectors.
Whether she plans to bring forward legislative proposals to regulate the use of artificial intelligence. Researchers at Robust Intelligence recently discovered that AI could bypass privacy protections within hours of use, raising concerns about sensitive personal data protection as Government's use of AI grows.
Our White Paper sets out how we will regulate AI through a flexible framework underpinned by five principles, including safety, transparency, explainability, fairness, accountability, governance, contestability, and redress. The Government will come back with proposals in the autumn following the White Paper consultation.
▸
Assessment & feedback
Response accuracy
Q2
Partial Answer
▸
Context
The question stems from the need for public trust in AI used by authorities, as well as the importance of consultation.
AI has been used by public authorities in various contexts affecting individual rights. Does the Secretary of State agree that public trust is essential for any changes to the law and that greater consultation is required?
The hon. Gentleman makes a thoughtful point. We need a comprehensive public debate, considering both risks and opportunities like job matching through technology. Further conversations will be had as we go forward with this critical technology.
▸
Assessment & feedback
Specific details on changes to the law or consultation plans were not provided.
Response accuracy
Q3
Partial Answer
▸
Context
The question is about protecting human rights through UK laws on AI, especially regarding collaboration with nations that abuse human rights.
Will the new UK laws on AI protect people by outlawing any collaboration or research with nations that seek to abuse human rights through facial recognition or social credit systems?
Specifically on AI technology, domestically we set out our proposals in the White Paper and internationally, we see a leadership opportunity for the United Kingdom as we are already a global leader in this technology and care passionately about ensuring its safe and responsible deployment.
▸
Assessment & feedback
Specific details on how UK laws will protect against human rights abuses were not addressed.
Leading The Charge Of Opportunity
Response accuracy
Q4
Direct Answer
▸
Context
The question is about public confidence in data protection as Government's use of AI expands.
Can the public have confidence that sensitive personal data will be protected given the expansion of automated decision making under the Data Protection and Digital Information (No. 2) Bill?
Yes, because the legislation focuses on protecting sensitive personal data. The White Paper sets out principles such as safety, transparency, explainability, fairness, accountability and governance, contestability, and redress to ensure protection.
▸
Assessment & feedback
Response accuracy
Q5
Partial Answer
▸
Context
The question is about confirming legislative plans given the rapid pace of AI deployment.
Will the Government confirm that they will bring forward legislation in the next Session regarding strengthening and clarifying regulators' mandates?
We said we anticipate introducing a statutory duty on regulators requiring them to have due regard to the principles. We will be returning to the House with an update on the conclusion of the White Paper consultation.
▸
Assessment & feedback
Specific timing and contents of legislation were not confirmed.
Cannot Commit To The Contents
Response accuracy
Q6
Direct Answer
▸
Context
The question is about taking an open and pragmatic approach in international AI diplomacy.
Does my right hon. Friend agree that an open, pragmatic approach needs to be shown to harness the benefits of AI for the economy and society?
My right hon. Friend is correct. We intend to lead overseas and domestically, leading the charge in public services and ensuring our pro-innovation approach enables capturing benefits sooner across the economy.
▸
Assessment & feedback
Response accuracy
Q7
Partial Answer
▸
Context
The question is about reassurance regarding comments made by Nick Clegg on AI's limitations.
Was my right hon. Friend reassured by the comments made by Nick Clegg this morning, representing Meta, that AI at present is not capable of individual thought?
We are drawing on the benefit of a range of experts in our work. I thank Ian Hogarth and those who have stepped up to form our Foundation Models Taskforce, which is looking closely at safety requirements for models at the very frontier.
▸
Assessment & feedback
Specific details about reassurance from Nick Clegg's comments were not provided.
Response accuracy
Q8
Partial Answer
▸
Context
Reports indicate bad actors using AI to create voice clones for scamming victims. McAfee found that one in four people had experienced a voice scam or knew someone who did.
The Secretary of State will surely have read the many reports of bad actors using AI technology to create voice clones and, ultimately, scam victims out of money over the phone. For as little as 76p a month, a subscription to one of those websites can be yours, Mr Speaker. Computer security company McAfee recently found that one in four people had experienced a voice scam or knew someone who had. What exactly is the Secretary of State doing right now to prevent and, ultimately, protect people from those voice scams and deepfakes?
The shadow Minister raises a good point, which we should all examine. We have the correct answers, through the approach we have set out in our White Paper using our world-class regulators, the approach we are working on in this House and the other place on the Online Safety Bill, and existing crimes and offences. Right now, my officials are working with Ofcom to ensure it is ready for the duties that will come to it through the Online Safety Bill, and that it has the tools to ensure that scams, fraud and other offences are properly addressed.
▸
Assessment & feedback
The Shadow Minister Raises A Good Point, Which We Should All Examine.
Response accuracy
Q9
Partial Answer
▸
Context
EU is moving towards legislating an AI Act to ensure beneficial use of AI technology. The questioner requests commitment from the Government.
The Secretary of State said that my colleagues had made thoughtful points about human rights concerns, control of personal data and public trust, but we need actual action and commitment from the Government. In contrast, the EU is moving to become the first regulator in the world to legislate for a specific AI Act, to ensure that AI works for people as well as business and Government. Will she finally commit to the UK following suit and legislating for AI, to ensure that the UK does not become a haven for the worst possible applications of what should be beneficial technology?
The hon. Gentleman is telling the wrong story. The UK Government are acting comprehensively. The framework we set out in our White Paper applies across the UK. The work we are setting out with our global summit leads the way internationally. That approach is proportionate and flexible, and has been welcomed by business across the UK. In addition, the principles I have already named and the work of my colleagues across Government on human rights and other aspects ensure that our citizens can enjoy the safe and responsible use of this technology. I look forward to working with anyone in this House who has a thoughtful approach to take to that.
▸
Assessment & feedback
The Hon. Gentleman Is Telling The Wrong Story.
Response accuracy