1 Viscount Waverley debates involving the Department for Science, Innovation & Technology

Advanced Artificial Intelligence

Viscount Waverley Excerpts
Monday 24th July 2023

(1 year, 4 months ago)

Lords Chamber
Read Full debate Read Hansard Text Watch Debate Read Debate Ministerial Extracts
Viscount Waverley Portrait Viscount Waverley (CB)
- View Speech - Hansard - -

My Lords, the takeaway from me from this informative debate is to ask: where do we go from here? Artificial intelligence is vital to our national interest, in regional prosperity, and for shared global challenges. It should be seen as an instrument that shows us the bigger picture in a vast, complex, and detailed chain over which no single country, or corporate, should ever have overall control. Caution must be assured that every data point, statistical analysis and prediction model be spot on. There may be dire consequences if we are ever reliant on data that is unverified, inaccurate or misleading due to misunderstandings of context or nuances that result in—in the jargon—technological hallucinations. Failing to do so could compromise the whole AI experience.

Got right, however, AI policy and regulation will play a vital role in monitoring compliance, analysing trends and assessing the impact of policies, providing transparency, trust and accountability that will ensure that AI-driven decisions and recommendations produce credible, far-reaching results. It can tell us, for example, where to seek proof of reliability, raise red flags and shed light on the previously invisible interconnection of the global economy by assisting us in understanding the complexities of trade dynamics.

Artificial intelligence is a comparatively unexplored component that can impact international trade. When used to help grow economies in a fast, efficient and fair manner, by pointing to the exact location of risks involved in a long-chain transaction, or a complex supply chain, this is all to the good. There is currently no system, however, that monitors and identifies suspicious global trade patterns, no mapping of complex international trade flows and no overall analysis of international trading patterns across multiple countries. AI can change all that. Nationally built systems in technological and political silos must be avoided to combat these challenges. However, collaborative efforts between nation states will enable a comprehensive understanding of patterns and devising targeted strategies. Collaborative efforts such as Project Perseus bring together technology, finance, and policy to unlock sustainable access for SMEs through data-sharing of accurate data from the real economy related to energy usage and resilience. This is critical for UK stakeholders in the business and banking world.

Some 200 million bills of lading, the documents at the heart of international trade, were recently examined by the International Centre for Trade Transparency and Monitoring, to which I am a party, revealing that 13.6% of these documents included at least one substantive inaccuracy. Such mistakes can quickly spread through a supply chain, posing risks that can have far-reaching effects on the economy, covering up inappropriate trade practices such as dumping, counterfeit or sanctions avoidance. Results that are 1 degree out skew detailed analysis.

ESG reporting is also becoming the new norm for companies to communicate their environmental, social and governance credentials to the markets. Interoperability affords legal protection and a process that safeguards SMEs and banks alike. AI also has great value in the prevention and detection of crime, especially fraud. When applied, it can save many staff hours and be a driver to get to the heart of the crime. What took a human researcher two days now takes an hour.

Before we launch into creating further uses of data, it is essential to ensure that government and industry have governance right. To underline all this, we need look no further than to a notice distributed to all Members in Parliament from the parliamentary digital department, informing us that

“generative AI tools are susceptible to bias and misinformation”.

We have the frameworks and processes in place to deliver success, but the time for theory is over.