YouTube secretly applies AI enhancements to Shorts videos

YouTube has confirmed it is using AI to enhance some videos on its Shorts platform without notifying content creators or asking for their consent.

Several creators, including Rick Beato and Rhett Shull, noticed subtle changes in their videos, such as altered skin texture and warped features, and raised concerns on social media about the undisclosed edits.

YouTube is running an experiment on select Shorts, using machine learning to unblur, denoise, and improve video clarity, similar to features found on modern smartphones.

The company did not clarify whether users will be able to opt out of these enhancements.

Experts, including Samuel Wooley from the University of Pittsburgh, pointed out that users can usually control such features on their own devices, while YouTube is making changes without creator permission.

The move has sparked debate about transparency and trust as AI continues to play a larger role in shaping online content.

🔗 Source: BBC

🧠 Food for thought

1️⃣ Undisclosed AI content erodes already fragile media trust

YouTube’s secret video enhancements highlight a critical transparency problem that directly damages public confidence in digital content.

Trust in mass media has already plummeted from 72% in the 1970s to just 34% in 2023, according to recent research1.

When platforms modify content without disclosure, this decline accelerates. Studies specifically show that not revealing AI-generated content negatively impacts audience trust2.

The YouTubers affected, like Rick Beato and Rhett Shull, expressed concern that these undisclosed changes could “erode the trust” they have with their audiences, reflecting broader anxiety about content authenticity.

This puts YouTube in a precarious position where their attempt to improve video quality may actually undermine the creator-audience relationships that drive their platform’s success.

2️⃣ Platform-level AI enhancement represents new frontier of content control

YouTube’s approach differs significantly from device-level AI features that users can control directly on their smartphones or cameras.