Meta will begin labeling AI-generated photos uploaded to its platforms like Facebook and Instagram over the next few months, as election seasons ramp up globally. The company will also start penalizing users who do not disclose if a realistic video or audio clip was created with AI.
According to Nick Clegg, Meta’s President of Global Affairs, these steps aim to “galvanize” the tech industry to address the growing issue of AI-generated content being indistinguishable from reality. Meta itself is developing tools to detect synthetic media even if metadata has been altered to hide AI’s role.
The company already watermarks images created with its Imagine AI generator as “Imagined by AI” and will now do the same for AI-generated photos made by Google, OpenAI, Microsoft, Adobe, Midjourney, and Shutterstock. However, Clegg admitted the industry lagged more on identifying AI video and audio. While vigilant about deception, Meta may still miss some things.
Meta to Require Disclosure of AI-Generated Content
The company will soon require users to disclose when sharing realistic AI video or audio, with penalties like post removal for non-compliance. Clegg believes it is “unlikely” that a completely synthetic but politically impactful video would evade early detection on its platforms.
The company has also begun internally testing large language models trained on its content policies to efficiently filter content for human review.
While this isn’t a miracle solution to counter the spread of misinformation using AI-generated content, it’s certainly a welcome step.
Emman has been writing technical and feature articles since 2010. Prior to this, he became one of the instructors at Asia Pacific College in 2008, and eventually landed a job as Business Analyst and Technical Writer at Integrated Open Source Solutions for almost 3 years.
