Defense Against Deepfakes

The advent of GenAI for image, voice, and video creation provided attackers with new tools to engage in scams and fraud. Brand damaging images are easy to create to tarnish the reputation of an executive or a company.

Protect2020, a UK consulting firm led by Andy Crocker, a former Metropolitan Police officer whose story was recounted by Joseph Menn in Fatal System Error, had this to say in a recently published white paper.

“For senior leaders and security teams, this is no longer a hypothetical future concern: deepfake crime is here, accelerating, and directly targeting corporate entities across the globe.”

Andy went on to describe the key characteristics of Deepfakes.

• Highly accurate voice cloning that can replicate tone, pitch, hesitations and emotional inflection.
• Video impersonation where entire face and head movements are synthesized to match the real individual.
• Synthetic identities created by blending real and fabricated features, enabling new forms of fraud.
• Low cost, low skill barrier, with publicly available tools allowing criminals to create realistic fakes in minutes.

With a new threat comes new defenses. These vendors are all addressing the Deepfake problem.