Meta Platforms (formerly known as Facebook) has announced plans to warn users about highly advanced videos and audio files from Artificial Intelligence that look deceptively real.

If users fail to do this, they face consequences, as Meta's Chief Policy Officer Nick Clegg emphasized on Tuesday in a blog post. The corporation plans to use warnings to label such content in its services.

This measure is necessary because it is not yet common to integrate invisible watermarks into AI software for videos and audio recordings, explained Clegg. However, he admitted that there are means and ways to remove these watermarks.

Meta is working on making this more difficult, though. The idea is to integrate the watermarks directly into the file creation process, which would make them impossible to remove. The company is also developing technologies to automatically detect AI-generated content.

Recently, automated calls in the USA caused a stir by imitating a deceptively real-sounding voice of President Joe Biden. The calls urged people not to participate in the Democratic Party's primary in the state of New Hampshire.

This incident has stoked concerns that in the coming months, attempts could be made to influence the outcome of the November presidential election using deceptively realistic AI forgeries.

As part of an industry alliance, Meta aims to establish technologies dedicated to uniquely labeling AI-generated files. The corporation has set a goal to tag AI images in its apps Facebook, Instagram, and Threads in all supported languages.

Additionally, Meta also offers AI software that creates images based on text prompts. These files contain both visible markings and invisible watermarks.

The stock of Meta Platforms temporarily fell by 0.79 percent to $455.76 on NASDAQ.