Meta Will Label AI-Generated Content In May

(CTN News) – Trying to reassure users and governments over the risks of deepfakes, Facebook and Instagram giant Meta announced on Friday that AI-generated media will be labeled starting in May.

Facebook has announced it will no longer remove manipulated images or audio that do not otherwise break its rules. Instead, it will rely on labeling and contextualization to ensure free speech is not violated.

Meta’s content moderation decisions are reviewed independently by its oversight board, which criticized the changes.

Based on the enormous advances in AI and the ease with which media can be manipulated into highly convincing deepfakes, the board requested Meta overhaul its approach to manipulated media in February.

In a pivotal election year not only in the United States but across the globe, the board sent its warning amid fears of rampant misuse of AI-powered applications.

Content such as video, audio, and images will be identified as “Made with AI” by Meta with its new labels. Content deemed to be at high risk of misleading the public will also receive a more prominent label.

A blog post by Monika Bickert, Meta’s Vice President of Content Policy, argues that providing transparency and context is the best way to address this content.

“The labels will include a broader scope of content than the manipulated content recommended by the Oversight Board,” she said. A February agreement between major technology companies and AI firms aims to crack down on manipulated content intended to deceive voters.

Meta, Google, and OpenAI have already agreed on a watermarking standard.

With AI-generated content labeling beginning in May 2024, Meta said removing manipulated media solely based on the old policy will cease in July. The new standard guarantees that content will remain on the platform, even if it is manipulated with AI, unless it violates other Community Standards, such as those prohibiting hate speech.

It has only heightened worries about the easily accessible technology after recent examples of convincing AI deepfakes have been discovered.

Last year, Meta left a manipulated video of US President Joe Biden online, prompting the board’s list of requests. It appears that Biden inappropriately touched the chest of his adult granddaughter when he voted with her. A robocall impersonating Biden urged tens of thousands of New Hampshire voters not to vote.

Artificial intelligence has been used to generate speeches for Pakistan’s Imran Khan party.

SEE ALSO:

ICTContact’s IVR Studio Unboxed, An In-Depth Look at Advanced Design Techniques