That's a really useful example. Thank you.
We do know, and there's plenty of evidence about this, that there is both deliberate and unconscious bias inherent in AI. I think there's quite a strong argument for a regulatory approach to govern AI deployment, much like we have in, say, the pharmaceutical sector. When you look at a product, before you can put it on the market, you have to look at what might be the unintended side effects of a particular medicine.
What do you feel about that? Do you think there is an argument for a regulatory approach, particularly because, as we know, the current deployment of AI does discriminate against women and does discriminate against black and ethnic minority citizens? People are losing jobs and are not gaining access to services like loans and mortgages because of this.