Deepfakes are getting faster than fact-checks, says digital forensics expert Hany Farid

image via New Scientist

February 17, 2026, 11:00 AM

• •Generative AI has made creating deepfakes faster, cheaper, and easier to scale.
• •The biggest misconception is that current AI is intelligence, it is just a 'token tumbler'.
• •Focus should be on going after infrastructure rather than individual content.
• •Liability needs to be created, and platforms that cause harm should be sued.
• •Public awareness and safety measures like safe words can minimize the impact of voice cloning scams.

Digital forensics expert Hany Farid discusses the challenges and potential solutions to deepfakes in the age of generative AI. He emphasizes the need to rebuild trust infrastructure by addressing the core issues of liability and the choke points that enable digital deception. Farid highlights the speed and scale of deepfake creation as a primary concern, emphasizing that the damage is often done before a fake can be disproven. He stresses that we must focus on the infrastructure that allows deepfakes to thrive to combat their negative impacts.

Read original article