Microsoft Bans DeepSeek App for Staff Over Security Risks

Microsoft Restricts Employee Use of DeepSeek App Amid Security and Propaganda Concerns In a significant move highlighting growing concerns over data security and foreign influence in AI, Microsoft's Vice Chair and President, Brad Smith, revealed during a Senate hearing on US AI capabilities that the company's employees are prohibited from using the DeepSeek app. The announcement, made on May 8th, underscores a cautious approach by the tech giant towards AI platforms with potential ties to foreign governments. Data Security and Propaganda: The Core Concerns The primary reasons cited for the internal ban on the DeepSeek app revolve around data security and the potential for propaganda. Smith articulated that Microsoft is wary of the risk that user data from DeepSeek could be stored in China and, consequently, be subject to Chinese law. This legal framework, as noted in reports, mandates cooperation with Chinese intelligence agencies. "At Microsoft, we don’t allow our employees to use the DeepSeek app," Smith stated, emphasizing the security risks associated with data being stored in China and potentially accessed by the Chinese government. Beyond data privacy, Microsoft is also concerned that DeepSeek’s responses to user queries might be influenced by "Chinese propaganda." A congressional report highlighted that the app "siphons data back to the People’s Republic of China (PRC), creates security vulnerabilities for its users, and relies on a model that covertly censors and manipulates information pursuant to Chinese law." DeepSeek's own privacy policy reportedly confirms that it stores data on Chinese servers, subject to Chinese legislation. Microsoft's Broader Stance on DeepSeek While the DeepSeek app is off-limits for Microsoft employees and is not offered in Microsoft's app store due to these concerns, the company's relationship with DeepSeek isn't entirely severed. Microsoft still offers the DeepSeek R1 model on its Azure cloud service. However, Smith revealed that Microsoft had taken steps to mitigate risks with the Azure-hosted model. He mentioned that the company had "managed to go inside DeepSeek’s AI model and ‘change’ it to remove ‘harmful side effects’." Furthermore, Microsoft stated that DeepSeek underwent "rigorous red teaming and safety evaluations" before its deployment on the Azure platform. The specifics of the modifications made to the model were not detailed. A Growing Trend of Restriction Microsoft is not alone in its cautious stance. The company joins a growing list of organizations and government bodies that have restricted or banned the use of DeepSeek. Several US government agencies, including the Department of Commerce and the US Navy, have already banned the AI model. Legislation such as the “No DeepSeek on Government Devices Act” could see these restrictions expand further. Internationally, organizations in countries like Australia, including the ABC, NBN, and AusPost, have also implemented similar bans for internal use. This public disclosure by Microsoft marks a significant moment, bringing to light the internal cybersecurity and geopolitical considerations major tech companies are grappling with as AI technology rapidly evolves and proliferates globally. The decision reflects a proactive approach to safeguarding sensitive corporate and user data and mitigating the potential spread of state-influenced information.

News

Trending

Today

This Week