Europe investigates big tech companies’ use of generative AI

CNN
—

The European Union on Thursday launched an investigation into Big Tech companies’ use of artificial intelligence and handling of computer-generated deepfakes, increasing scrutiny as technology officials fear they could disrupt elections.

The study covers companies such as Meta, Microsoft, Snap, TikTok, and Focusing on something.

“The European Commission recommends that these services implement measures to mitigate the risks associated with generative AI, including so-called “hallucinations” in which AI provides false information, the viral spread of deepfakes, and automated manipulation of data. We are asking you to provide us with further information. It is a service that has the potential to mislead voters,” officials said in a release.

European Commission regulators have said they are particularly concerned about the potential for generated AI to cause disruption in the run-up to this summer’s EU parliamentary elections. Online platforms have until April 5 to answer questions about steps they have taken to prevent AI tools from spreading election misinformation.

“We are asking our platforms whether they are prepared for a kind of 11th-hour injection scenario right before an election, where high-impact deepfakes could be distributed at scale, and how prepared they are for this kind of scenario. What’s going on?” commission officials told reporters on a conference call Thursday.

Part of the committee’s goal is to gain insight into how companies are tackling the issue of deepfakes, and to ensure that AI-related incidents are not included in the landmark technology regulation, the Digital Services Act. It is also a reminder to businesses that they may be subject to fines and other penalties based on the Act. Laws governing social media and other major online platforms.

The companies’ responses could be incorporated into a set of election security guidelines for high-tech platforms that the European Commission is expected to finalize by March 27, another commission official said.

The AI ​​research also covered broader topics, including how platforms are addressing the impact of generative AI on user privacy, intellectual property, civil rights, and child safety and mental health. ing.

Companies have until April 26 to submit their answers to these questions.

The request for information sent to X this week is related to an ongoing investigation into Elon Musk’s social media company that began shortly after the outbreak of the Israel-Hamas conflict last year, officials said.

“One of the complaints we have is that the service can be manipulated through automated means, which may include generative AI, so yes, there is relevance to ongoing investigations. ” said one committee official.

X CEO Linda Yaccarino met with Thierry Breton, head of the EU digital regulator, in late February.