Microsoft and OpenAI are investigating whether a group associated with the Chinese AI startup DeepSeek improperly accessed data from OpenAI’s systems, according to a report by Bloomberg.
Microsoft’s security researchers detected individuals believed to be linked to DeepSeek extracting large amounts of data via OpenAI’s API. Under OpenAI’s terms of service, developers must pay to use its API, which imposes strict limits on the amount of data they can access.
As OpenAI’s largest investor, Microsoft flagged the activity as a potential violation of OpenAI’s terms of service.
DeepSeek’s AI Breakthrough Raises Concerns
DeepSeek’s latest AI model, R1, has reportedly outperformed U.S. rivals at a fraction of the cost, triggering a selloff that wiped out $1 trillion in market value across global tech stocks.
David Sacks, an advisor to former U.S. President Donald Trump, told Fox News that there is substantial evidence suggesting DeepSeek used a technique called “distillation” to extract knowledge from OpenAI models. He further suggested that U.S. AI firms will likely take action to prevent such techniques, which could slow down competing AI models.
Despite these concerns, Sacks maintained that the U.S. remains ahead in AI but acknowledged that DeepSeek is closing the gap rapidly.
OpenAI and Microsoft React
OpenAI stated that Chinese firms are continually attempting to extract knowledge from U.S. AI models and emphasized its ongoing collaboration with the U.S. government to protect its intellectual property.
However, OpenAI did not respond to specific requests for comment on DeepSeek. A company spokesperson acknowledged the ongoing risks posed by AI competitors, stating:
“As the leading builder of AI, we are working closely with the U.S. government to ensure that the most capable AI models remain protected from adversaries and competitors.”
Microsoft also declined to comment on the ongoing investigation.
Global Impact of DeepSeek’s AI Model
The discovery of potential data breaches has raised serious concerns about data security and intellectual property (IP) protection, with possible implications for both companies and the broader AI industry.
The controversy follows DeepSeek’s recent release of its open-source AI model, R1, earlier this month. DeepSeek-R1 has quickly become the centre of attention in the global AI market. Its launch contributed to a significant decline in global technology stocks, reportedly shaking investor confidence in the AI sector’s demand for high-tech chips.
DeepSeek-R1 claims to be an improved reasoning model that outperforms current standards on several crucial AI tasks. However, like other Chinese-made AI models, DeepSeek self-censors on politically sensitive topics in China.
Unlike ChatGPT, DeepSeek deflects questions about Tiananmen Square, President Xi Jinping, or the possibility of China invading Taiwan.
The Future of AI Data Security
The implications of this potential data breach investigation could have far-reaching effects on the AI industry. The outcome of the probe will play a crucial role in shaping the future of AI data security and privacy.
As AI technology continues to evolve, companies like Microsoft, OpenAI, and DeepSeek will be under increasing scrutiny regarding intellectual property protection, ethical AI development, and data security.