AI image generator Midjourney will be blocking users from creating deepfakes of US President Joe Biden and Donald Trump as the 2024 US Presidential Elections loom closer.
Midjourney CEO David Holz announced on Wednesday that it is time to "put some foots down on election-related stuff for a bit" amid a surge of political disinformation fueled by AI, Associated Press reported.
(Photo : Morry Gash - Pool/Getty Images)
Holz admitted that "this moderation stuff is kind of hard" for an AI platform hosting 1.2 million to 2.5 million users daily.
Midjourney cited only 11 staff, including its CEO, on its website working for the company.
As of writing, the names of the two presidential candidates will immediately be flagged as an "abuse alert."
According to reports, the clampdown on the images related to political figures will only be temporary as it prepares for further policy changes.
Also Read: AI Image Generators are Spreading Election Misinformation, New Study Finds
AI Disinformation Surges Ahead of 2024 Elections
The policy change was reported just a few days after watchdog group Center for Countering Digital Hate published a study noting how popular AI chatbots still generate political misinformation.
Among the mentioned AI generators were Midjourney, ChatGPT, and DreamStudio.
In a statement to CNN, Holz vouched that "updates related specifically to the upcoming US election are coming soon."
Several AI firms, including OpenAI and Anthropic, have earlier joined collaborative efforts with other tech giants to strengthen their guardrails against potential abuse of their AI models during the election period.
So far, many chatbots have been reported to prevent users from requesting image generation related to the 2024 Elections, although some were still able to find some loopholes.
Related Article: Google Restricts Any Election-Related Queries for Gemini Chatbot in U.S., India
Still No AI Law in the US
Despite the growing danger posed by the misuse of the technology to the people and democracy, the US is still struggling to pass a law for a standardized regulation on AI.
Biden issued an executive order last October to address safety concerns for the technology but has left most planning and preparation to the promised regulatory move to individual agencies.
On the other hand, Trump has labeled the recent AI innovations as "dangerous and scary" but did not provide further plans to mitigate its risks.
It is worth noting that prior to his impeachment, Trump had signed an executive order allowing the employment of AI technology within government offices.
