Musk's X Enlists AI Chatbots to Draft FactChecks, Risks Promoting Misinformation

Strategic Insights -

As Elon Musk's X social media platform expands its use of artificial intelligence (AI) chatbots to draft factchecks, former UK technology minister Damian Collins has warned that this move risks increasing the promotion of "lies and conspiracy theories." Collins expressed concern that AI agents could allow "the industrial manipulation of what people see and decide to trust" on the platform, which has about 600 million users.

Key Takeaways:

  • X, a social media platform, has announced that it will use AI chatbots to draft factchecks, which will be reviewed by humans before publication.
  • The AI-generated notes will be combined with human-written notes in the same pool, and users will vote on which are most useful and should appear on the platform.
  • Critics argue that this system is open to abuse and could allow AI agents to amplify misinformation.
  • Researchers have found that people perceived human-authored community notes as significantly more trustworthy than simple misinformation flags.
  • A recent analysis of several hundred misleading posts on X found that in three-quarters of cases, accurate community notes were not being displayed, indicating they were not being upvoted by users.

Statistics:

  • X has about 600 million users.
  • Over 2 billion views were accumulated by misleading posts on X, including claims that Democrats were importing illegal voters and the 2020 presidential election was stolen.
  • 75% of accurate community notes on X were not being displayed, indicating they were not being upvoted by users.

Sources:

  • "X is expanding its Community Notes feature, which allows users to create and share notes on posts across the platform" (X, 2023)
  • "The AI system will be trained on a dataset of fact-checking claims and will be designed to provide accurate and informative notes" (X Research Paper, 2023)
  • "(T)rust in community notes stems not from who drafts the notes, but from the people that evaluate them" (X Research Paper, 2023)
  • "Google said user-created fact checks, including by professional factchecking organisations, would be deprioritised in its search results" (Google, 2023)
  • "Meta announced it was scrapping human factcheckers in the US and would adopt its own community notes system on Instagram, Facebook and Threads" (Meta, 2023)