Synthesia AI video generator

Transparency and accountability 🙈

Synthesia is seen to suffer from a number of important transparency and accountability limitations.

• Algorithmic black box. The AI models and algorithms used are proprietary, with limited public information on their inner workings.

• Attribution. It can be difficult to determine if a video generated using Synthesia technology was AI-generated or to trace its origin.

• Auditability. The process of creating AI videos is not easily auditable by third parties.

• Legal accountability. The lack of industry-wide standards or regulations governing AI video generation means it is unclear who is responsible if AI-generated content causes harm or spreads misinformation.

Risks and harms 🛑

Synthesia has been criticised for the misuse of its AI-powered stock videos, in particular, to create political misinformation and disinformation, and to facilitate fraud and scams. 

The ease with which Synthesia avatars can be misused has also psychologically damaged the actors whose likeneness were created and negatively impacted their career opportunities.

Legal, regulatory 👩🏼‍⚖️

• Synthesia response to Ofcom call for evidence (pdf)

News, commentary, analysis 🗞️

• https://spectrum.ieee.org/deepfake-cottage-industry

• https://futurism.com/the-byte/deepfakes-translation-tech-synthesia

• https://www.wired.com/story/synthesia-ai-deepfakes-it-control-riparbelli/