AI-powered software can disinform people in elections, NGO report warns

Generative AI tools such as Midjourney and ChatGPT are being criticized for producing deceptive images of political candidates, fueling concerns about election disinformation, TRTWorld cited a report published on Wednesday.

A recent report by the Center for Countering Digital Hate (CCDH) revealed that these AI image tools generated false content in 41% of cases, with Midjourney being the worst performer at 65%.

It added that tech giants, including Meta and Microsoft, are uniting to combat misleading AI content using technologies like invisible watermarks.

The article also cited an engineer at Microsoft raising alarms over AI image generators like DALL-E 3 and Copilot Designer, stating that they can produce harmful content, including sexual objectification and political bias.

The report emphasized the need to prevent such content from causing harm to communities and democracy.