OpenAI, Google will watermark AI-generated content to hinder deepfakes, misinfo

Seven companies promised Biden they would take concrete steps to enhance AI safety.

Read in full here:

This thread was posted by one of our members via one of our news source trackers.

Watermarking is currently easy to foil, by making minor (and easy-to-automate) changes to text. I haven’t yet seen any reasearch showing a way to reliably watermark text that is 90% generated and 10% edited.

1 Like