The bottom line: Deepfakes are being created faster than they can be identified — traditional visual verification is no longer reliable enough for the speed of today’s AI-generated content.
Leading deepfake researcher Hany Farid warns that traditional methods for authenticating visual content cannot keep pace with the speed of modern AI generation procedures. This presents organizations with significant challenges in verifying authenticity before manipulated content spreads virally.
Hany Farid, recognized expert in deepfake detection, documents the growing dilemma between technical forgery capability and verification possibilities. The development of generative AI models enables the synthesis of audiovisual content with increasing quality, while established analysis and validation methods are reaching their limits.
For Chief Data Officers and security professionals, this concretely means: The previous assumption that experts can reliably identify manipulated media through visual inspection or technical artifact analysis is increasingly no longer valid. The speed at which synthetic content can be produced outpaces the capacity for manual or semi-automated verification in real time.
The problem is not only technical but also organizational: viral spread occurs within hours, while authenticity analyses can take days. This fundamentally requires new strategies for data verification, provenance tracking, and transparent labeling of synthetic content at the point of creation, rather than relying solely on post-hoc detection.
Source: www.golem.de · Published 22 June 2026
Lumi AI News — AI-assisted curation pursuant to Art. 50 EU AI Act. Paraphrasing and classification by Lumi News Pipeline v1.7.1.