Nvidia-backed AI company Synthesia on Thursday (25 April) unveiled a new line of AI-generated digital avatars that are able to convey human emotions through user prompts.

Synthesia said its AI ‘expressive avatars’ eliminate the need for traditional recording equipment and costs for professional video production. The London-based company said it trains the AI with actors in front of a green screen in a studio. 

The AI tool is currently used by over 55,000 businesses to deploy digital avatars for training videos and corporate presentations, according to the company.

In a recent video demonstration released by the company, a user inputs, “I am very happy. I am so sad. I am frustrated”, to which the AI-generated avatar reads out in the corresponding emotion.

Synthesia was founded in 2017 and raised $90m from investors in 2023, valuing it at $1bn. 

In order to fight misuse of the technology, such as spreading fake news and deepfake avatars, the company said all users must sign up as enterprise customers in order to create synthetic avatars.  

Synthesia said that all of the content made with its technology is moderated by its team. 

The UK company also required its new customers to undergo a thorough vetting process, much like that in the banking industry. Synthesia said this stops any bad actors from gaining access to the technology for misuse.

GlobalData forecasts that the overall AI market will be worth $909bn by 2030, registering a compound annual growth rate (GAGR) of 35% between 2022 and 2030.

In the GenAI space, revenues are expected to grow from $1.8bn in 2022 to $33bn in 2027 at a CAGR of 80%.