YouTube, the video streaming platform owned by Google’s parent company Alphabet, has released new guidelines on labelling AI-generated and synthetic content.

The platform will now require creators to disclose to viewers when realistic content – defined as content a viewer could easily mistake for a real person, place, or event – is made with altered or synthetic media, including generative AI.

Users can accomplish this using a new tool in YouTube’s Creator Studio, and the disclosures will appear as labels in the expanded description or on the front of the video player.

“Generative AI is transforming the ways creators express themselves – from storyboarding ideas to experimenting with tools that enhance the creative process,” YouTube said is a statement. “But viewers increasingly want more transparency about whether the content they’re seeing is altered or synthetic.”

YouTube is not requiring creators to disclose content that it deems as “clearly” unrealistic, animated, includes special effects, or has used generative AI for production assistance.

Examples of content that will require an AI label include: using the likeness of a realistic person, altering footage of real events or places, or generating realistic scenes.

The move comes after YouTube announced in November that it would require users to flag AI-generated content.

Just a week ago, Google announced that it would crack down on AI-generated content transparency in an effort to safeguard the upcoming elections in 2024 – their AI chatbot, Gemini, will not answer questions relating to the elections taking place in India, and will also restrict answers regarding elections in the UK and the US taking place later this year.

YouTube will not require users to label content that it deems clearly unrealistic, nor will creators have to flag adjustments to colour or light, special effects like background blur, or beauty filters.

For videos with more sensitive subjects – like news, health, elections, or finance – a more prominent label will be shown if it contains synthetically-generated material that needs to be flagged under the policy.

YouTube says that these tools and labels will be rolled out in the coming weeks, starting with mobile apps.

Recommended reading

• ICO Will Call on G7 Countries to Tackle Cookie Pop-ups
• UK Gov Announces £100 Million for AI Regulation & Research
• Tech Giants Work on Deepfake Pact Ahead of Election Year

The company also says they are working on a system for people to request the removal of AI-generated or altered content which simulates an identifiable individual.

Since the rise of advanced AI, the public has been understandably concerned about the rise of disinformation and the use of deepfakes – altered videos and images that using people’s likeness to realistically depict them in fake scenarios.

Tech giants, including Microsoft, Meta, Google, TikTok, and OpenAI have entered negotiations to curtail the rise of deepfakes amid the upcoming elections.

The World Economic Forum claimed that misinformation and disinformation – largely powered by AI – is the world’s top short-term threat.

With 969 million registered voters in India alone, 2024’s elections can dramatically shift global politics, and it appears that companies are starting to understand the potential fallout of unregulated advanced AI.