Metropolitan Digital

Google


.

Watermarking ChatGPT, DALL-E and other generative AIs could help protect against fraud and misinformation

  • Written by Hany Farid, Professor of Computer Science, University of California, Berkeley
Watermarking ChatGPT, DALL-E and other generative AIs could help protect against fraud and misinformationImages generated by AI systems, like these fake photos of Donald Trump being arrested (he hasn't been arrested), can be a dangerous source of misinformation.AP Photo/J. David Ake

Shortly after rumors leaked of former President Donald Trump’s impending indictment, images purporting to show his arrest appeared online. These images looked like...

Read more: Watermarking ChatGPT, DALL-E and other generative AIs could help protect against fraud and...

More Articles ...

  1. How To Outsourcing Your Support System?
  2. Israel's military reservists are joining protests – potentially transforming a political crisis into a security crisis
  3. Reaction to bronze sculpture of Coretta and Martin Luther King Jr. in Boston hasn't been good – and that's not bad for art that shatters conventions
  4. 40 years ago 'A Nation at Risk' warned of a 'rising tide of mediocrity' in US schools – has anything changed?
  5. How do superconductors work? A physicist explains what it means to have resistance-free electricity
  6. 3D-printing the brain's blood vessels with silicone could improve and personalize neurosurgery – new technique shows how
  7. Public radio can help solve the local news crisis -- but that would require expanding staff and coverage
  8. Federal Reserve’s ‘soft landing’ goal has become bumpier with rate hike plan hit by bank turbulence
  9. Back to the Moon: A space lawyer and planetary scientist on what it will take to share the benefits of new lunar exploration - Podcast
  10. Infant formula shortages forced some parents to feed their babies in less healthy ways