YouTube will soon introduce AI-powered tools to shorts. This allows users to generate videos using a simple text prompt. Remove camera needs and editing tools. In his keynote address at the Cannes Lions 2025 Festival, the company's CEO, Neal Mohan, announced an update that uses Google's latest video generation model, Veo 3.
“We've already seen this on YouTube,” Mohan said. “VEO is a video generation model for Google Deepmind. This allows you to create video clips for AI-generated backgrounds and shorts.” Future versions will go further, allowing you to create not only the background but the entire video.
With YouTube shorts already taking in over 200 billion daily views, this new feature raises some immediate questions.
What is YouTube adding?
Veo 3 is Allows users to generate the complete video clip Visuals and Audio – Based on text prompts. This is a major upgrade from the current Dream Screen feature, only allowing users to create AI backgrounds.
YouTube says the new tool will be rolled out in the coming months, but it has not yet said whether it will be limited to paid AI Pro or AI Ultra plans.
How does this affect content creators?
YouTube shorts play a major role in the platform, promoting billions of daily viewing and providing a key revenue stream for many content creators.
This dynamic can be confusing as AI-powered video generation is now available to everyone. As a result, YouTube algorithms can start to surface more AI-generated content, as users can generate faster and larger scale. Creators who rely on monetizing their shorts could soon find themselves competing with the waves of AI-made videos.
Mohan described VEO 3 as having “a new creative lane for everyone to explore.” But content creators may wonder: at what cost?
Copyright and liability
There are also legal questions. Although Indian copyright law does not mention AI, AI-generated videos that mimic famous characters, movies or songs without permission can infringe the rights of the original creator.
The use of AI does not exempt anyone who generates or uploads such content from responsibility. Recent cases show that the courts are already dealing with content such as copyright infringement.
In August 2024, the Delhi High Court ruled that the Deepfake video of the character of Taarak Mehta Ka Ooltah Chashmah violated both copyright and trademark protection and issued an injunction against such content.
If similar AI-generated shorts appear on YouTube, users will likely bear primary legal liability. Platforms like YouTube are protected under India's Intermediate Safe Harbor Act (Section 79 of the IT Act) as long as they act promptly to remove infringing content after receiving actual knowledge, such as formal notices and court orders. The legal liability of AI model providers such as Google is still a grey region in India, and there is no major case law yet to clarify this.
It is also important to note that Indian copyright laws do not currently recognize AI as an author. Section 2(d) of the Copyright Act defines “author” as a natural person. In early 2020, the Copyright Bureau rejected attempts to register artwork created solely by the AI tool “Raghav”, and subsequently put the registration on hold after the addition of human co-authors.
As a result, works generated purely by AI may lack clear copyright ownership in India. However, if AI-generated shorts incorporate protected elements of existing films, music, or characters, they could infringe the original creator's rights under sections 13 and 14 of the Copyright Act, which defines protected works and exclusive rights.
How does YouTube ease AI shorts?
YouTube previously announced that content creators should disclose if their videos contain realistic AI-generated content. These videos will display public labels as they are known to the audience.
The platform has also updated its Privacy and Impersonation Policy to allow individuals to request takedowns of content that misuses likeness and voice.
advertisement
YouTube's content ID system remains the primary method of copyright detection. However, while effective for direct use of copyrighted material, content IDs can sometimes be hard to detect AI-generated videos that mimic styles and characters without using accurate copying.
Regarding the Veo 3 itself, Mohan said the model “has a significant improvement in video quality and incorporates audio”, but it also includes filters designed to block prompts to generate copyrighted content or accurate replicas of famous people. How these filters work consistently is an open question.
So far, noteworthy cases
Recent cases show how quickly this legitimate space is evolving.
Global trends
Discussions about AI-generated content and copyright are unfolding globally.
EU AI law requires AI model providers to immediately disclose whether they use copyrighted material in their training and prevent illegal output. Article 17 of the EU Copyright Directive also strengthens the platform's obligation to prevent infringement.
In the US, the Copyright Office issued guidance that if human authors are not sufficient, AI-generated works cannot be protected by copyright.
Also, when it comes to AI, India is watching closely, as seen in the parliamentary committee report, which flags the gap in current copyright law. In April 2025, the government also established a new DPIIT-led committee to study whether Indian copyright laws can adequately address AI-generated content and new legal questions related to it.
Why is it important?
By making AI video generation widely available within YouTube shorts, Google is trying to reconstruct the landscape of short content. Anyone can create videos quickly and easily without the skills and tools of a human creator.
For the YouTube creator community, this raises real questions.
- Will independent content creators lose revenue or visibility if AI-generated shorts start to dominate their user feeds?
- And how effectively can the platform mitigate this flood of new content due to unresolved legal and ethical issues regarding copyright impersonation?
- Do AI-generated shorts start to crowd content created by human creators?
As Mohan said, “The possibilities for AI are endless.” However, it is still unclear whether it is good news for content creators or for rights holders trying to protect their work.
Medianama will query Google for these issues and update the story if it receives a response.
Read again:
Supporting our journalism:
for you
Source link
