German courts rule Google liable for what its AI search says (madrobot.blog)

A German court has ruled that Google is liable for false claims generated by its AI search feature, a precedent that raises significant implications for the AI/ML community. In a case involving a doctor in Aachen, the court found that Google’s AI erroneously attributed a criminal conviction to him, which belonged to a different individual. The court emphasized that Google must take responsibility for AI-generated content, asserting that failing to verify the accuracy of such outputs could lead to legal consequences, including fines of up to €250,000 for each breach. This ruling follows a similar decision from the Munich Regional Court, which also determined that AI-generated summaries constitute independent statements for which Google is accountable. This denotes a shift in legal responsibility, suggesting that AI responses are not mere reflections of other sources, but rather generate unique content that carries the same obligations as traditional media. With these decisions, Google may face increased scrutiny and regulation regarding its AI functionalities in Europe, while contrasting legal outcomes in other jurisdictions, like the US, highlight the ongoing debate about the role of AI in media and information dissemination.