Meta Tags AI Content Across Platforms

Meta is implementing new ways to detect and label AI-generated images across its platforms to promote transparency around this emerging technology.

Summary

  • Meta has been working with industry partners to develop common technical standards for identifying AI-generated content through invisible markers like metadata and watermarks.
  • In the coming months, Meta will start labeling AI images across Facebook, Instagram, and Threads when its systems detect these common industry indicators.
  • Meta will require users to disclose if they are sharing realistic AI-generated video/audio and apply labels in those cases.
  • The company is also developing classifiers to detect AI content lacking markers and ways to prevent alteration/removal of watermarks.
  • Meta notes challenges in identifying all AI content currently and that adversarial efforts to deceive may increase.
  • The company emphasizes its community standards apply regardless of content origin and is testing using AI to improve policy enforcement.
  • Meta aims to be transparent about AI content while developing new generative AI responsibly.

READ ARTICLE

Related post

AI Ethics

DOJ to Punish AI-Enabled Crimes More HarshlyTags:

US Justice Department is directing federal prosecutors to pursue harsher penalties against criminals who have used AI to facilitate or advance their misconduct. There is a particular focus on election security and misuse of AI around the 2024 elections. READ MORE

AI moderation

Meta Expands AI Image Labels to Curb Election Misinfo

Meta will expand efforts to identify and label AI-generated images on its platforms ahead of upcoming elections globally, seeking to curb misinformation and deception. Meta will label AI-generated images from major AI companies like Google, OpenAI, Microsoft etc. in addition to content created by its own AI tools. These labels…