YouTube to Strengthen Deepfake Detection in 2026 as Gemini API Requests Double

GOOGGOOG

YouTube CEO Neal Mohan said YouTube will deploy enhanced deepfake detection and expand likeness detection to millions of creators in 2026 to curb low-quality AI-generated content. Google Cloud’s Gemini API requests more than doubled between March and August, while Google-backed Isomorphic Labs delayed its first clinical trials until late 2026.

1. YouTube Sets AI Content Integrity Goals for 2026

In his annual letter, YouTube CEO Neal Mohan identified reducing “AI slop” and detecting deepfakes as top priorities for next year. The company will expand its existing spam- and clickbait-fighting systems to curb the flood of low-quality AI-generated videos, require creators to clearly label AI-produced or altered content, and roll out enhanced likeness detection to millions of Partner Program creators. These measures build on YouTube’s removal of harmful synthetic media and follow the December launch of deepfake face-flagging tools that aim to protect creator identities and maintain advertiser trust.

2. Alphabet’s AI Ecosystem Fuels Margin Expansion and Growth

Alphabet reported that its services segment—comprising search advertising, YouTube, subscriptions and devices—delivered a 40% operating margin through the first nine months of 2025, while its cloud division achieved a 21% margin. The company’s vertical integration strategy leverages DeepMind research, custom tensor processing units and global data-center investments to lower AI training and inference costs. The recent $4.7 billion acquisition of an energy infrastructure business will further stabilize power supply for its AI workloads. This end-to-end approach positions Alphabet to capitalize on continued revenue acceleration and profit expansion as enterprises ramp up AI deployments.

Sources

PSFCZ
+5 more