A warning that AI-generated fakes are eroding trust in what people see and hear, which is bad for anyone trying to tell real from fake online.

• AI can now fake video and audio well enough that you can't trust your own eyes for the first time in history.
• The problem has shifted from finding information to verifying it — the hard part is knowing if what you see is real.
• This trust breakdown opens the door to large-scale misinformation and manipulation of what people believe.
• The World Economic Forum has flagged fake and misleading information as one of the biggest global risks.

Outlook: Expect the fight over what's real to intensify as AI fakes spread faster than tools to catch them.