In 2026, deepfakes will no longer be novel; they will be routine, scalable, and cheap, blurring the line between the real and the fake.
Hany Farid
On Berkeley News↗ ·
ArticleonAI
WhatSayAI take
AI-assisted editorial framing · not reportingWHATSAYAIUSEFUL
CROWDSPLIT
The line is strong because it does not argue that deepfakes are coming; it argues they are becoming normalized infrastructure. That makes the real issue less novelty than scale, cost, and the collapse of trust.
Editor's note
Strong because it turns the deepfake debate into a trust-at-scale problem.
Crowd verdict
How readers judged this take
Split verdict
Vote distribution
0primary votes
- Accurate0(0%)
- Useful0(0%)
- Speculative0(0%)
- Hype0(0%)
Needs sources · 0 flags
Live reader tally.
The context
The quote appears in UC Berkeley’s roundup of AI expert predictions for 2026, focused on the growing social and democratic risks of increasingly convincing AI-generated media.
What's your take?
Tap to register your view.
Pass along