ENISA, the EU cybersecurity agency, has warned that powerful new AI models might become a disruptive factor in the EU elections next June as malicious actors could use them to run large-scale information manipulation campaigns.
The warning comes as part of the 2023 Threat Landscape, in which the cybersecurity agency takes stock of the evolving trends annually. The spotlight is on possible disruptions related to AI chatbots, like the renowned ChatGPT and any form of AI-fuelled information manipulation.
“Trust in the EU electoral process will critically depend on our capacity to rely on cybersecure infrastructures and on the integrity and availability of information. Now, it is up to us to ensure we take the necessary actions to achieve this sensitive yet essential goal for our democracies,” ENISA’s executive director, Juhan Lepassaar, said in a statement.
The fear that powerful new AI models could be used to disrupt electoral processes is not unique to Europe. With elections coming up in the United States and the United Kingdom, experts have warned about the risk of AI-made images or videos going viral and about AI-powered propaganda bots.
AI models are particularly relevant for psychological manipulation, so-called social engineering, thanks to their capacity to produce human-like text or authentically-looking images and videos replacing one person’s likeness with another, known as deepfakes.
ENISA’s report notes that, since the public release of ChatGPT [the early demo version in November 2022], Artificial Intelligence has been the main innovation driver in social engineering, notably thanks to the technology’s capacity to produce text that mimics a legitimate source, and do voice cloning and AI-powered data mining.
In March, Euractiv reported that another EU agency, Europol, in charge of law enforcement, had pointed out the potential criminal uses of generative AI systems like ChatGPT for online fraud and other cybercrimes.
Europol warns against potential criminal uses for ChatGPT and the likes
The EU law enforcement agency published a flash report on Monday (27 March) warning that ChatGPT and other generative AI systems can be employed for online fraud and other cybercrimes.
ENISA wrote that “the trend suggests that generative AI is providing an avenue for threat actors to craft sophisticated and targeted attacks at speed and scale”, adding that it anticipates “more targeted social engineering attacks using AI-based technology in the future”.
Voice cloning, the process of synthetically reproducing a person’s voice with only a few seconds of videos posted on social media, is mentioned as a growing threat. For instance, it could be used to fake the voice of a family member and ask for a ransom, or to denigrate a politician publicly.
“The disruptive adoption of generative AI and AI chatbots is rapidly changing the threat landscape, where the ability to detect AI-generated content or AI-based interaction becomes a matter of urgency,” the report warned.
AI-powered chatbots like OpenAI’s ChatGPT and Google’s Bard might also be subject to manipulation of the datasets, for instance, to put a business or political competitor in a bad light in the system’s responses.
Meanwhile, EU policymakers have been finalising the AI Act, the world’s first comprehensive regulation on Artificial Intelligence. However, how to deal with generative AI and deepfakes is still very much an open discussion.
AI Act: EU countries headed to tiered approach on foundation models amid broader compromise
The EU approach to powerful AI models is taking shape as European countries discuss possible concessions in the upcoming negotiations on the world’s first comprehensive Artificial Intelligence (AI) rulebook.
The issue of deepfakes has come to the fore in Spain, the country currently holding the rotating presidency of the EU Council, after the image manipulation technique was used to produce faked naked pictures of teenage girls.
The first test of a deepfake tilting an election occurred earlier this year in Turkey, where one of the main candidates for the presidential elections withdrew from the race after a manipulated sex tape went viral.
[Edited by Zoran Radosavljevic]
Read more with EURACTIV
Latvia gets front row in rebuilding Ukraine’s telecom infrastructureThe Latvian government committed on Wednesday (18 October) to help Ukraine restore and develop its telecom infrastructures, support it in meeting the EU telecom acquis, and help it to participate in EU and international programmes.
Source: euractiv.com
