We must fundamentally change the way we think about Artificial Intelligence


The Science, Innovation, and Technology Committee has released its final report for the 2019-24 Parliament, concluding its inquiry into the governance of Artificial Intelligence (AI). This report examines both domestic and international developments in AI governance and regulation since the Committee’s interim report in August 2023.

The report emphasizes that its conclusions and recommendations are relevant to the upcoming government post-General Election. While affirming the current sectoral approach to regulation, the Committee suggests that the next government should be prepared to enact legislation on AI if any regulatory gaps emerge in addressing public interests in this rapidly evolving field.

Revisiting the Twelve Challenges of AI Governance outlined in the interim report, the Committee provides recommendations for policymakers on how to address these challenges. One of the most significant challenges identified is the opacity of AI systems, often referred to as a ‘black box’, where the basis and reasoning behind AI outputs may be obscure, yet their predictive capabilities can surpass human abilities. In response, the Committee advocates for stronger testing of AI outputs to evaluate their effectiveness and accuracy.

Expressing concern over reports that the new AI Safety Institute has encountered difficulties accessing some developers’ models for pre-deployment safety testing, the Committee urges the next government to identify and publicly name any developers who refuse such access. This aligns with the agreement made at the November 2023 Summit at Bletchley Park. The Committee stresses the importance of equipping UK regulators with adequate resources to hold AI developers accountable, citing the disparity between regulatory capabilities and the vast resources available to developers.

Chair of the Science, Innovation, and Technology Committee, Rt Hon Greg Clark MP, highlights the need for a shift in assessing AI technology due to the inherent challenges posed by ‘black box’ models. He underscores the importance of testing AI outputs to detect any biases or undesirable consequences. Additionally, he calls on the next government to be proactive in addressing regulatory shortcomings and ensuring that regulators have the necessary statutory powers to be effective in overseeing AI developments.

In conclusion, the Committee emphasizes the transformative potential of AI while emphasizing the importance of safeguarding public protections. It urges the new government to follow the agenda outlined in the report to harness the benefits of AI responsibly.

Source: committees.parliament.uk