Google forced to delay Bard AI’s EU launch over privacy concerns

Europeans wanting to try Google Bard will have to wait. The Irish Data Protection Commission (IDPC), the main overseer of data in the European Union, has forced Google to delay the rollout of its Bard chatbot in the region. The generative AI was supposed to launch in the EU this week, but IDPC Deputy Commissioner Graham Doyle says his agency hasn’t received a “detailed” privacy briefing, a data impact assessment or supporting info.

The Commission is still in the midst of an “ongoing examination” of Bard, according to Doyle. It isn’t estimating when it might wrap up that investigation, but it plans to share info with other EU data regulators as quickly as possible.

We’ve asked Google for comment. In a statement to Politico, a Google spokesperson says the company promised to expand Bard access “responsibly” after discussing its efforts with experts and governments. This includes talking to privacy regulars to “address their questions and hear feedback,” the representative says.

Google began widening access to Bard in March, when it let would-be users sign up for a waiting list. In May, it dropped the waitlist and expanded availability to 180 more countries and territories. The tech giant shied away from the EU, where the General Data Protection Regulation (GDPR) sets strict requirements on how personal data is collected and shared.

Other AI developers have already faced tighter scrutiny. Germany, Italy and Spain are currently investigating OpenAI’s ChatGPT alongside the European Data Protection Board, and Italy temporarily banned the tool over worries both the chatbot and its training methods (using others’ real-world content) might violate the GDPR. A delay for Bard theoretically minimizes the chances of similar conflicts for Google.