Scope: AI

Can't find what you're looking for? Try the search bar!

Synthetic data: accuracy and real-life limitations

This does not surprise me – as I was explaining yesterday to two podcast hosts, synthetic data is great on paper but it does not help with accuracy or real-life outliers/divergent situations. It’s easy to claim that synthetic data is the best way to comply with data minimisation and that “you don’t really need all […]

Read Analysis →

2025 reflections and looking ahead to 2026 trends

With the festive season upon us, I may be posting a little less and instead singing Christmas songs with the family. So let’s look back at 2025 and anticipate 2026, based on interactions with my LinkedIn articles & posts: – Pseudonymisation & the concept of “personal data” were significant in 2025. 4 of my top […]

Read Analysis →

What the EDPB AI opinion and Italian OpenAI fines mean for data models

Myth-busting time! This week, the EDPB published its Opinion on AI models & personal data, and then the Italian authority revealed it had fined OpenAI. But unlike what some are suggesting, the Italian decision *predates* the EDPB Opinion by a month. So let’s dispel a few myths surrounding that Opinion: – “Game changer”? The positions […]

Read Analysis →

My first thoughts on the EDPB opinion regarding AI models and personal data

Full analysis likely to come in the coming days, but some first thoughts on the EDPB’s “AI models & personal data” Opinion. In essence, the EDPB doesn’t exclude anything entirely but leaves it up to national supervisory authorities to make a call in cases they are handling – and while it doesn’t prohibit relying on […]

Read Analysis →

GDPR Digital Omnibus debate: key takeaways

Do (re)watch our 53-minute civilised, proper debate on some of the key GDPR parts of the Digital Omnibus proposal! From the notion of “personal data” to special categories of data, AI development & training and even scientific research, a lot of ground covered from two perspectives. [I had a few awkward moments of the “unmute” […]

Read Analysis →

Analysis of the GDPR and Data Act Omnibus draft

From fishing (instead of the GDPR) to a random “v” having appeared instead of a few words (a typo upon deletion?), the Commission’s GDPR & Data Act Omnibus feels like a rushed document. But what does it cover really? I’m focussing here on GDPR & ePrivacy: – Personal data definition: I like the idea of […]

Read Analysis →

Nordic Privacy Summit: balancing SCD definitions and workability

Hopefully thought-provoking? My keynote on “Personal Data in a Digital World: What Are We Protecting?” kicked off the Nordic Privacy & Innovation Summit 2025 today in Stockholm. Among my topics: SRB, OT, ePrivacy, AI and AI model training, plus how to make the GDPR more workable (see slides, which show that data protection involves a […]

Read Analysis →

Navigating the leaked GDPR and Data Act Omnibus document

GDPR & Data Act Digital Omnibus: this is *NOT* the final version we expect the Commission to publish in two weeks (already made public by some outlets such as Contexte). But this way anyone keen on reading in more detail the points I summarised and commented in my previous post on the GDPR aspects (see […]

Read Analysis →

Detailed breakdown of key changes in the GDPR Omnibus draft

Key points from (& thoughts on) draft GDPR Digital Omnibus: First, Art. 4(1) GDPR would be completed with an SRB-like statement (great news), plus a mitigation of abusive interpretation of the Scania judgment (also great news). Art. 4(15) on “data concerning health” would also finally be more specific, avoiding some of the more awkward and […]

Read Analysis →

Overview and preliminary thoughts on the GDPR Omnibus draft

That draft GDPR Digital Omnibus is quite worth the read. Once the story gets out, I’ll be posting a summary of all the changes and some preliminary thoughts – but it’s a promising blend of data protection case law (+ mitigations of excessive interpretations that might be made of certain judgments), ePrivacy improvements and AI […]

Read Analysis →