Blue AI symbol against a blurred digital background

Photo by Zach M / Unsplash

AI video generation is lowering barriers to entry in the gig economy

Freelancer earnings for using AI-generated video and video editing skills on Upwork increased by 329% year-on-year in 2025, according to Upwork's annual In-Demand Skills 2026 report, despite the generative AI marketplace squeezing freelance incomes in other categories. AI video generation is lowering the barriers to entry that have traditionally excluded a new video freelancer from the gig economy. In contrast, while freelance video editing is on the rise, almost all freelance jobs in text-based occupations have lost ground due to the introduction of generative AI

One reason for the differences between video and other genres of work is a recent study published by Xiang Hui, Oren Reshef, and Luofeng Zhou in their article, "The Short-Term Effects of Generative Artificial Intelligence on Employment". The study indicates that the number of job postings and the associated earnings of freelancers who supplied text-based and image-based work have declined dramatically since the launch of generative AI models — most notably, top-rated freelancers, who are typically the most highly skilled. In contrast, video freelance jobs are growing faster than all other AI skill categories in the report, according to Upwork's own 2026 skills data report, which indicates that AI video generation and video editing is now the fastest-growing of Upwork's skill categories, followed by AI integration, data annotation, and chatbot development. The stark contrast between the rapid growth of AI video generation and editing as compared to training for text-based and image-based jobs illustrates that while generative AI has revolutionised the way people create image-based and text-based work, it did not do so to the same degree as it did with respect to video production. Prior to the introduction of generative AI tools, video production was accompanied by a much higher entry barrier, as opposed to text-based work, for example, where low-cost software existed on every laptop. Video production has always had high barriers to entry due to the required equipment, cost of materials, and years of training required to be considered qualified. Often this counted as a disadvantage for new entrants into video production.

With the introduction of Seedance 2.5 and similar tools, most of these barriers to entry have come crashing down. These new technologies produce 4K high-definition motion video (less than 30 seconds in length) from simple text descriptions or a small number of reference images. In addition, with Seedance-style tools, there is no need for equipment such as camera rigs or cranes because you can now create virtual camera movement (dolly and crane motions, etc.) with your editing software. Finally, an editor no longer has to wait until they receive footage back from the original source before they are able to edit something in a particular area. In the past, if a freelancer with a laptop had no filmmaking background or training, it would have been virtually impossible for them to produce the quality of footage required for a professional video.

To illustrate how far we've come, imagine that a single seller on Etsy hired a freelance videographer to produce a video advertisement for their product. They would pay several hundred dollars for a 30-second ad, and then wait on the videographer's schedule to complete it. Once the ad was finished, the seller would have to pay again if the lighting was incorrect. Today, that same seller could generate a similar video in an afternoon and pay a freelancer just to provide direction and fine-tune the final product. The same shift can be seen from the labour supply side: with the advent of AI-assisted editing, a video editor can service many more clients than they previously could because the slow part of the editing process (shooting and assembling) no longer falls under the video editor's responsibility.

The lack of challenge in editing does not mean it is economically less relevant. The demand for editing services stems from the fact that companies want finished videos, and not from the standpoint that editing itself is highly valued. Since a machine can create raw footage for a video, the only valuable skill a freelance editor can bring is their judgement (i.e., knowing which cuts, takes, and prompts are most appropriate for a specific project). Research conducted by Upwork substantiates this with a statement that businesses are now hiring people to produce and modify AI-generated content, rather than people to operate cameras.

Additionally, since most of the video editing marketplace is controlled by a small number of platforms, these companies are able to use monopsony power over freelancers, meaning they can set the prices and terms of business. With lower barriers to entry in video production, the quantity of people working at entry-level positions is expected to increase; consequently, competition for lower-skilled jobs will increase, thereby placing downward pressure on pay rates. While lower barriers to entry make the market more accessible, they don't necessarily improve workers' prospects. Instead, they create a more competitive market, increasing pressure on freelancers to secure work.