The Legality of a Possible Copyright Lawsuit against DeepSeek
- Sabrina Soto
- Apr 11
- 2 min read
Naomi Moneme
In January 2025, DeepSeek, a Chinese AI startup, announced that it had developed an AI Model, R1. What sets DeepSeek apart from its competitors is the cost factor. Their relatively low development costs for an AI app are around $6 million, while OpenAI and Google investments total $1 trillion. Despite its low investment costs, DeepSeek has developed an efficient model with superior learning abilities and memory utilization, which allows it to outperform its competitors.
At Microsoft, researchers believe that DeepSeek has allegedly been withdrawing large amounts of its data using OpenAI’s models. This was done by individuals who were hired at Microsoft and had connections with DeepSeek to remove large amounts of data. A spokesperson for Microsoft, “We take aggressive, proactive countermeasures to protect our technology and will continue working closely with the US government to protect the most capable models being built here.” In just a short time DeepSeek has been able to surpass ChatGPT in downloads and set a new stock price for tech companies in the United States.
DeepSeek's accusations of utilizing OpenAI’s models to train their AI model, R1. have led to questions about Intellectual Property (IP) rights and what are the legalities that exist within the AI industry. OpenAI’s ChatGPT is currency protected under copyright law, they argue that since AI is the content it is not a human and is not legally a person. This case differs from typical Copyright cases because with AI there are more nuances involved in the case.
In the defense of DeepSeek, there is an argument that the data they used for their R1 model used Fair Use, which means that under the legal principle that allows for some copyrighted material in special circumstances such as teaching or criticizing the material without payment or permission from the holder of the information.
Deepseek could claim that ChatGPT's outputs train its model. Ap possess that involves using publicly available data to build more efficient systems. DeepSeek could claim that by using ChatGPt’s output to train its model to fall under fair use under the “Transformative Use”, which means that they are using ChatGPT’s information in a separate and new context from the original purpose. In this specific case, DeepSeek is taking this data that is available to the public to build a farther more efficient AI model.
Sources
Clarence-Smith, Louisa and Sellman, Mark. (29 January 2024). OpenAI investigates possible data theft after DeepSeek debut. The Times
Edmond, Charlotte. (5 February, 2025). What is open-source AI and how could DeepSeek change the industry? Work Economic Forum
Zilber, Ariel and Barrabi, Thomas. (29 January 2024). OpenAI says it has proof DeepSeek used its technology to develop its AI model. NYpost.
Madigan, Kevin. (29 May 2023). Landmark Warhol Decision Reins in Transformative Fair Use. Copyright Alliance.
