Italy’s Scorza: publicly available data is not for training algorithms

“Privacy rights are becoming commercial and technological assets which have an enormous impact on markets and geopolitics,” Guido Scorza, a Board Member of Italy’s Data Protection Authority, the Garante, said.

Speaking in his personal capacity at the Privacy Laws & Business 37th International Conference yesterday, Scorza said that data that is publicly available online is not to be used to train algorithms: “At least data subjects must have an easy and effective way to opt-out. This will be difficult in practice, but we must try.”

He said that the Garante will continue its investigation into Open AI, even if temporarily stopping its processing was an unpopular decision for Italian consumers. While not wishing to stop innovation, Scorza said that the Garante will try to act as fast as possible on AI questions. However, sometimes this means making mistakes. He also stressed the role of education.

“The right to do business and to innovate are both fundamental rights - we must avoid putting one in front of the other,” he concluded.

Mid-June, Meta announced that it had paused its plan to use posts and photos from Facebook and Instagram to train its AI models. The news follows action by some EU and UK Data Protection Authorities, and complaints by noyb.

The PL&B conference continues today and tomorrow.