The EU on Thursday wielded a powerful new digital law to press TikTok, Facebook and six other platforms on the risk of AI to elections, including through deepfakes.
In a flurry of actions taken under the EU’s Digital Services Act (DSA), the European Commission quizzed TikTok, Facebook, Instagram, X, Google, YouTube, Snapchat and Bing on what they were doing to counter those risks.
The commission also announced a formal probe against Chinese internet retailer AliExpress for suspected multiple breaches of the DSA, among them the sale of illegal medicines and dietary supplements, and not preventing minors from accessing pornography.
Additionally, Brussels asked Microsoft’s professional social network LinkedIn about how users’ personal information is being used for targeted advertising.
- Budget padding: I warned Ningi but he didn’t listen – Ndume
- NIGERIA DAILY: Ways To Build A Resilient Nigeria
On the information request to the eight platforms on steps to mitigate the use of generative AI, the commission said in a statement it was looking at risks “such as so-called ‘hallucinations’ where AI provides false information, the viral dissemination of deepfakes, as well as the automated manipulation of services that can mislead voters”.
The formal requests for information made to those platforms, as well as the request made to LinkedIn, do not presuppose further action being taken. That would depend on the information gleaned.
The formal probe against AliExpress, on the other hand, gives Brussels the power to dig deep into the platform’s internal documents, and processes and to take testimony.
Such a level of action against a company, depending on the outcome, potentially exposes it to DSA fines running up to six per cent of a platform’s global turnover, or even a ban in egregious cases.
European Commission officials said the focus on generative AI and how major platforms were handling it stemmed from concerns about how it might be used to influence voting in the June EU elections.
“We want to equip ourselves and we want to equip the platforms and alert the platforms to really be best prepared for all sorts of incidents that might come our way with regard to the upcoming elections, in particular, of course, the EP (European Parliament) election,” one official told journalists.
The announcement looking at the AI risk to EU elections came a day after the European Parliament voted to adopt a new law that seeks to curb abuses of artificial intelligence. That legislation will take effect once formally signed off on by EU member countries.
