|
Good morning, Marketers, and here’s a problem for ChatGPT.
I think it was generally agreed among experts at last week’s MarTech that there would be teething problems with generative AI and that, while it’s here to stay, developing it to its full potential would hit some bumps in the road.
It hit a massive bump on Friday. It turns out that thoughtful legal minds have been pointing out for a while that large language models of this kind might be in breach of GDPR in a number of ways, especially because the data they are trained on likely includes personal data as defined by the regulation. It’s hard, for example, to see how ChatGPT could meaningfully respect the right to be forgotten.
Looks like this is all going to be tested sooner than expected, with Italy imposing a preemptive ban while asking OpenAI to consider moving towards compliance. That means that, as of today, ChatGPT cannot process the personal data of any Italians. And it’s unlikely to stop there.
Kim Davis,
Editorial Director
|
bit.ly/mtnl
