AI View April 2024

Welcome to AI View, Simmons & Simmons’ fortnightly round-up of key AI legislative, regulatory, and policy updates from around the world.

This edition brings you:

• Proposed UK AI Bill receives second reading in House of Lords
• IPPR report finds that 8 million UK jobs are at risk from AI
• UK & US announce partnership on AI safety
• MEPs endorse bill for AI regulation in the workplace
• UK Cabinet Office issues Guidance on AI Procurement for Public Sector

Proposed UK AI Bill receives second reading in House of Lords

On 22 March, the UK AI (Regulation) Bill received its second reading in the House of Lords. This Bill originated as a Private Members’ Bill, meaning that it is ultimately unlikely to become law.

If passed, the Bill will establish an AI Authority, which would have various responsibilities, including:

• coordinating with regulators to ensure they take account of AI in fulfilling their duties;
• pinpointing any gaps in the AI regulatory landscape; and
• ensuring a consistent approach across industries, rather than a “piecemeal” approach if AI regulation is left to individual regulators.

The proposed Bill would also require organisations to be open about their AI practices and ensure that these practices are in line with relevant consumer, data protection, and intellectual property laws.

The Bill received broad support during its second reading. However, there are concerns about whether the Government will adopt the Bill, with some Lords expressing opposition to creating another regulatory authority, in trying to balance a ‘pro-innovation’ stance to the adoption of AI.

Read the proposed UK AI (Regulation) Bill here.

IPPR report finds that 8 million UK jobs are at risk from AI

The Institute for Public Policy Research (IPPR) has released a report warning that up to 8 million jobs in the UK could be at risk from AI without government regulation. Back office, entry-level and part-time positions are particularly vulnerable.

The report outlines two waves of AI adoption. The first, described as here and now, is said to be putting 11% of workers’ routine tasks at risk, and the second wave, where companies will integrate AI more deeply, could affect 59% of tasks, including more specialised tasks in higher-earning jobs.

The IPPR has modelled three scenarios for the second wave of AI adoption, dependent on certain policy choices. These range from a ‘worst case’ scenario with significant job losses and no GDP gains, to a ‘best case’ scenario with no job losses and huge GDP growth:

• Worst case scenario – full displacement: all at-risk jobs are replaced by AI, no GDP gains and 7.9m jobs lost.
• Central scenario: 4.4m jobs lost but with economic gains of 6.3% of GDP (£144bn a year).
• Best case scenario – full augmentation: all jobs at risk are augmented to adapt to AI instead of being replaced, leading to no job losses and growth of 13% GDP (£206bn per year).

The IPPR urges the government to adopt a job-centric industrial strategy to manage the transition, promote job transitions and ensure the benefits of automation are widely shared.

Key measures include:

• Supporting green jobs that are less exposed to automation.
• Tax incentives and subsidies to encourage job-augmentation rather than displacement.
• Regulation to ensure human oversight over key issues, such as health.

Read the full report here.

UK & US announce partnership on AI safety

On 1 April, the UK and US signed a Memorandum of Understanding to collaborate on AI safety research, evaluations and guidance.

The agreement aims to align scientific approaches and accelerate the development of robust AI evaluation methods.

The UK and US AI Safety Institutes will share information, cooperate closely, and consider expert personnel exchanges to build shared capabilities. They plan to conduct joint testing exercises on publicly accessible AI models and create a common approach to AI safety testing.

Read the full announcement here.

MEPs endorse bill for AI regulation in the workplace

The European Parliament’s Employment and Social Affairs Committee has endorsed the new Platform Work Directive, which aims to introduce new rules to improve the working conditions of people engaged in ‘platform work’: i.e., where an online platform is used for the provision of specific solutions or services (otherwise referred to as the gig economy). The Directive aims to help determine the correct employment status of platform workers and to regulate the use of algorithmic systems and AI in relation to platform work.

These digital platforms often use algorithms and automated decision-making systems to oversee, manage, and assess the performance of workers on the platform. The Directive will ensure that any person performing platform work cannot be dismissed or fired based on a decision taken by an algorithm, AI or an automated decision-making system. Instead, relevant platforms must ensure human oversight on important decisions that directly affect the persons performing platform work.

The next step is for the EU Parliament to formally endorse the Directive in April, before it is adopted by the Council. EU Member States will have two years after publication in the Official Journal to integrate the Directive into their national legislation.

Find out more here.

UK Cabinet Office Issues Guidance on AI Procurement for Public Sector

On 26 March, the UK Cabinet Office produced a Policy Note addressing the growing use of AI in government procurement. The Policy Note provides guidance regarding risk management and the appropriate use of AI within public services.

The guidance outlines that public sector bodies must:

• evaluate potential risks and opportunities that come with AI adoption;
• require suppliers to disclose their use of AI when submitting bids and undertaking due diligence on their AI usage; and
• take measures to ensure confidential or non-public information is not used as training data for AI systems when preparing bids for tenders.

Read the Policy Note here.