The Deepfake Revolution of 2025

I just came across a very interesting article published on The Conversation. In it, Dr. Siwei Lyu, a famous computer scientist and an IEEE Fellow, explains why 2025 was the “breakout year” for synthetic media and why 2026 is likely to become even more unusual.

Based on the article, here are the highlights of what’s changing:

• Perfect Realism: New AI models have solved the "glitches," creating videos with perfect consistency and voices that sound 100% human.

• Zero Barriers: High-quality fakes no longer require a tech genius—anyone with a chatbot can create them in minutes.

• The "Real-Time" Future: We are moving toward a world where live video calls can be faked in real-time, making "behavioral coherence" the next big battleground.

As the line between "real" and "synthetic" disappears, Dr. Lyu argues that we can no longer rely on our own senses. Instead, we need to lean on new digital watermarking and forensic tools to verify the truth.

Here is the full article.