Expert Thoughts on the UK’s AI Safety Summit
November 2023 by Experts
Today the UK AI Safety Summit begins in Bletchley Park. Experts have offered their insights and hopes for the outcome of the summit below, including Ori Bendet of Checkmarx, who points out that accountability is key and Paul Brucciani of WithSecure, who points out a key consideration for AI regulators will be how AI may affect our lives in the future.
Ori Bendet VP of Product Management at Checkmarx: Accountability is key
“The executive order announced by the US President will put new responsibilities on the companies that are developing AI systems and the focus going forward needs to be on the accountability that we all have in reducing risks. As the UK prepares for its own AI Safety Summit, it presents an opportunity to highlight the real challenges that the AI era holds for everyone, from technology companies to consumers.
With the rapid pace of development, everyone needs to be educated. One of the topics billed for discussion is what AI developers should do to scale their technology responsibly and I’d like to see those attending providing actionable and operational steps that vendors should take in order to reduce the risks presented by AI solutions.“
Paul Brucciani, Cybersecurity Advisor at WithSecure: How will AI affect our lives?
“It’s easy to be freaked out by the sci-fi imaginings of the destruction that singularity will bring, but how likely are they? The recent warnings by the US Centre for AI Safety (CAIS) are of apocalyptic impact but are they probable? Realising these risks requires: finance; technological challenges to be overcome; and the collusion of third parties. The table below assesses the gravest threats identified by CAIS. Each one has at least one significant obstacle to it being realised.
A table with different colored squares Description automatically generated
It may be that we overestimate the risk of an AI-driven, dystopian future, but AI will be ubiquitous, so we should consider how it will affect our lives. AI regulators are debating this question: the European Commission has draft regulations to minimise AI’s detrimental effect and the White House has proposed human rights principles that may find their way into federal law. There is some distance between EU and US positions. It a global agreement is reached on the use of AI, it won’t be soon.”