Italian Court Annuls €15M ChatGPT Privacy Fine
A Rome court ruled Italy's Garante had no jurisdiction once OpenAI set up its Irish subsidiary, cancelling the fine without ruling on the merits.
A Rome court has annulled Italy's €15 million GDPR fine against OpenAI, PPC Land reports.
The original fine
In November 2024, Italy's data protection authority, the Garante, fined OpenAI €15 million and ordered a six-month public awareness campaign across radio, TV, newspapers and the internet. It alleged failures including not reporting a March 2023 data breach, lacking a legal basis for training ChatGPT, transparency problems and weak age checks.
The Garante's formal decision, numbered Provvedimento n. 755 and dated November 2, 2024, accused OpenAI of violating several specific GDPR articles: Article 33 for failing to notify the Garante about the data breach of March 20, 2023; Articles 5(2) and 6 for allegedly failing to establish a lawful basis for processing personal data before training ChatGPT's models; Articles 5(1)(a), 12 and 13 over transparency shortcomings in its privacy policy; and Articles 24 and 25(1) for not implementing adequate age-verification systems Tribunale di Roma judgment, via Sapienza University.
Why the court cancelled it
The court ruled on a procedural point: once OpenAI set up its Irish subsidiary in 2024, Ireland's regulator became its lead authority under the GDPR's "one-stop shop" system. The Italian Garante therefore had no jurisdiction to issue the fine.
The court did not decide whether OpenAI actually broke the GDPR. Reuters reported that the Rome court's ruling, issued on March 19, 2026, cancelled the fine without immediately releasing its full reasoning Reuters. OpenAI welcomed the decision in a statement carried by Reuters, saying it has "always been committed to respecting users' privacy" and looks forward to helping more people, businesses and Italian society benefit from AI, according to Italian outlet *la Repubblica*'s translation of the company's comment la Repubblica.
When the court's full written reasoning became available months later, in judgment number 4153/2026, legal analysts confirmed that the court had annulled the Garante's decision "in its entirety on a purely preliminary, jurisdictional ground," finding that the regulator "simply lacked the competence to adopt the decision at all." Every other ground of appeal — including arguments going to the substance of the GDPR violations and to the size of the penalty — was declared "absorbed" and never examined by the court European Law Blog.
Why it matters
The ruling means complaints about big AI companies with an EU headquarters will mostly be handled by the country where they are based — often Ireland. That centralises enforcement but has long been criticised for slowing it down.
Legal commentators have described the underlying 2024 Garante decision as "the only final GDPR enforcement decision ever adopted in Europe in relation to the public launch of generative AI" at the time it was issued, underlining how significant its subsequent annulment is for the broader question of how Europe regulates AI companies European Law Blog.
Background: Italy and ChatGPT
Italy's Garante was the first Western regulator to take action against ChatGPT. In March 2023 it temporarily banned the service in Italy over privacy concerns, after a bug exposed some users' chat titles and payment details. OpenAI restored access a few weeks later after adding age checks, privacy notices and ways for users to object to their data being used.
The €15 million fine in late 2024 was the conclusion of that investigation, and it was widely seen as a signal that European regulators were willing to tackle generative AI directly.
How the one-stop shop works
Under the GDPR, a company with a main establishment in the EU is mainly supervised by the regulator in that country — its "lead supervisory authority". Other national regulators pass complaints to the lead authority and can only act alone in urgent cases.
Many large tech companies have their EU headquarters in Ireland, which is why Ireland's Data Protection Commission handles cases against several of the world's biggest platforms. Critics argue this concentrates a huge workload on one regulator; supporters say it stops companies facing 27 conflicting decisions.
A detailed analysis from the DPO Workspace legal newsletter summarized the court's position succinctly: "the issue is not the merits but jurisdiction," noting that for cross-border processing like ChatGPT's, GDPR enforcement power sits with the lead authority — in this case, Ireland's Data Protection Commission — rather than with the regulator of any individual member state where the company's users happen to be located DPO Workspace.
What happens now
The ruling does not clear OpenAI. Concerns about training data, transparency and age checks can still be examined through the Irish regulator and the European Data Protection Board, which has published guidance on how GDPR applies to AI models. The Garante may also appeal.
As the DPO Workspace analysis put it, the ruling "does not certify ChatGPT's GDPR compliance: it only decides who had power to sanction," meaning the substantive obligations on companies that use or build on ChatGPT remain unchanged DPO Workspace.
The case also exposes what some legal scholars have called an "AI enforcement gap": because OpenAI only recently established its Irish entity, there was a window in which no single regulator had clear, exercised authority over its earlier conduct in Europe, since the Garante's jurisdiction evaporated once the Irish establishment was recognized as lead authority but Ireland's own enforcement timeline for the underlying conduct remains separate European Law Blog.
What it means for you
If you are in the EU and want to object to how an AI company uses your data, you can still complain to your own national regulator. It will pass the complaint on to the lead authority where needed.
What readers should do
- If you use ChatGPT or other AI tools and have privacy concerns, you can still lodge a complaint with your own country's data protection authority — it does not need to be filed directly with Ireland's regulator.
- Review your AI chat history and account settings periodically; most major AI providers, including OpenAI, offer settings to control whether your conversations are used for model training.
- Be aware that a company's EU headquarters location (often Ireland) determines which regulator leads on cross-border enforcement, which can affect how quickly a complaint is investigated.
- Follow announcements from the European Data Protection Board for EU-wide guidance on how GDPR applies specifically to generative AI systems, since this area of law is still developing.
- If you run a business processing EU users' data through AI tools, understand that jurisdictional rulings like this one do not remove your own GDPR obligations around legal basis, transparency and data subject rights.
Sources & Verification
Additional reporting
- PPC Landppc.land
- Reutersreuters.com
- la Repubblicarepubblica.it
- European Law Blogeuropeanlawblog.eu
- DPO Workspacedpoworkspace.eu
- Tribunale di Roma judgment, via Sapienza Universitydei.web.uniroma1.it



