EU: Content created by AI and capable of spreading fake news must be marked

As part of the fight against fake news, content created by artificial intelligence and capable of spreading disinformation must be marked, said Vera Jourová, the member of the European Commission responsible for transparency and values.

At her press conference in Brussels, Vera Jourová emphasized that it should be the task of companies operating systems using artificial intelligence to label the generated content in order to inform users as part of the package of measures aimed at combating fake news and disinformation from Russia. Companies that integrate artificial intelligence into their services must build safeguards into their systems so that the services cannot be used for harmful purposes, he said. “Companies operating services created by artificial intelligence capable of spreading fake news must introduce technology that recognizes such content and flags them for the information of users,” he said.

They want the user to be able to “clearly see” that the content is not created by real people and be tagged with words like “the robot is talking here”. According to Jourová, social media companies should regulate the potential “dark side” of artificial intelligence, which can fake or create events and sounds in seconds. Ten days ago, he met with Google CEO Sundar Pichai and asked him if his company had technology that could detect fake news. “His answer was: Yes, but we are still developing our technology,” he said.

As part of a broader effort to combat Russian disinformation, the EU has also asked Facebook and others to devote more resources to fact-checking content in minority languages ​​and to Eastern Europe, where Russian disinformation campaigns are seen as a threat. “This is not part of the normal course of business. The Russians want to achieve this by undermining public support among our citizens who support Ukraine,” said Vera Jourová. “We simply have to protect our interests, our democracy; we have to protect – I can say this – our struggle and the war as well, because what we are doing is supporting the Ukrainians’ claim to win the war”.

According to Jourová, companies – including Google, Microsoft and Meta – that have joined the EU’s code of conduct to combat disinformation must report in July on safeguards in place to deal with disinformation potentially spread by artificial intelligence. “I have said many times that the protection of freedom of speech is our main task, but when it comes to the production of content created by artificial intelligence, I do not see that machines have the right to freedom of speech”. The European Commission wants the AI ​​nomination system to be up and running as soon as possible.

Last week, Twitter terminated its cooperation under the relevant EU code, and the vice president warned the American community site that it could expect regulatory scrutiny in the near future. “By abandoning the code, Twitter attracted a lot of attention and chose confrontation. We believe that this was a mistake on Twitter’s part. Twitter chose the more difficult path,” said the EU commissioner, adding that they will firmly and urgently check that the whether the American company’s activities comply with the EU regulations on new digital services that will enter into force on August 25. Launched in 2018, the voluntary code has been joined by 44 online platforms, with Twitter being the only major platform to withdraw so far. Many believe the commission will not hesitate to make an example of Twitter to show the DSA has teeth.