Zoom’s new AI avatars raise deepfake concerns

Concerns are growing over Zoom’s AI avatars, with fears they could be used for deepfakes.

Zoom plans to launch AI avatars by 2025, allowing users to create digital clones to improve workplace communication.

Zoom plans to roll out custom AI avatars by 2025, offering users a photorealistic digital clone that can replicate head and arm movements. Users will be able to script what their avatar says, with audio synced to lip movements, making it a potential game-changer for asynchronous workplace communication.

The company sees these avatars as a time-saving tool, particularly for those needing to scale video content creation. Despite the promising benefits, questions remain about how Zoom will prevent misuse of this technology, particularly in light of growing concerns around deepfakes. Many fear these digital avatars could be used maliciously to impersonate others or spread disinformation.

Zoom has announced some safety measures, such as advanced authentication and watermarks, though details remain unclear. Other companies offering similar tools, such as Tavus and Microsoft, have stricter safeguards in place, raising concerns that Zoom’s efforts may not be sufficient to curb potential risks.

The rise of deepfakes continues to fuel regulatory debates. While some states in the US have introduced laws to combat the spread of AI-aided impersonation, there is still no comprehensive federal regulation addressing the issue. As Zoom prepares to launch this feature, the company will need to balance innovation with security.