Good to know the regulator is tackling the "hard GDPR question" of accuracy in generative AI head-on. Part 4 of the generative AI consultation: We don't cover the right to rectification here, we did it in Part 3. Part 3: We don't cover rectification here, we'll do it in Part 4.

The situation with OpenAI and Scarlett Johansson is someone asking ChatGPT to write a 2024 version of the 1988 Midler v Ford case. Spoiler alert: Midler won. en.wikipedia.org/wiki/Midler_v.….

I'm sure the ICO had good reasons not to enforce against Snap, but I struggle to see its investigation as a "warning shot". Snap continued to offer its "My AI" feature throughout the investigation. There's no fine, reprimand, or actual warning afaik So what's the warning shot?

The ICO has updated its generative AI consultation based on my somewhat snarky tweet. If anyone know how to reconcile a controller's obligations under the GDPR's "right to rectification" and the inherent unreliabilty of Large Language Models, please respond to it (I don't).

Apple has not "killed" the various companies people are saying it has. Spotify exists despite Apple Music. Netflix persists despite Apple TV. Not to mention Android, Chrome, Linux, Firefox, Brave, etc. 1Password etc will remain a better option for many people and enterprises.

A glimpse inside the weird world of Community Notes.

When listening to Tony Blair reel off dubious predictions about the revolutionary impact of AI, remember that this man reportedly only started using a computer in the final year of his third term as Prime Minister.

The CJEU has ironed out a seriously obscure ambiguity around when civil society groups can bring litigation on behalf of data subjects curia.europa.eu/juris/document… It reminded me of one of my best Twitter threads—from two years ago, when I still wrote those—about a very similar case

CPPA proposes to finally fix this mistake in the CCPA Regulations, thus PROVING ME RIGHT when I said it was obviously a mistake. If you're disclosing personal information to a third party, then—by definition—you aren't doing so for business purposes. Take THAT, doubters.

Having a bad morning? Apparently this is worth a try. Disclaimer: I have no clue whether this is a good idea, but some people I respect appear to approve of it.

Whenever #databreach is trending, X usually just presents a huge list of posts about seemingly unrelated data breaches That's not due to a design flaw, it's because serious data breaches are happening everywhere, all the time I feel like we're officially past the tipping point

I don't get it.