India Warns Government Officials Against Using AI Tools like ChatGPT and DeepSeek Over Security Concerns

WHAT HAS HAPPENED?

The Ministry of Finance has issued a directive to government officials, advising them to “strictly avoid” using Artificial Intelligence (AI) tools, including ChatGPT and DeepSeek, on official devices. The warning comes due to concerns over the confidentiality of government data and documents. DeepSeek, developed by a Chinese company, was launched in January 2025. The circular, accessed by The Hindu, was issued by the Department of Expenditure (DoE) on January 29, 2025. It warns that these AI tools pose a risk to sensitive government information.

The circular states, “It has been determined that AI tools and AI Apps (such as ChatGPT, DeepSeek, etc.) in the office computers and devices pose risks for confidentiality of Government data and documents. Therefore, the use of AI tools/AI Apps in office devices may be strictly avoided.” The directive was issued by Pradeep Kumar Singh, Joint Secretary at DoE, with the approval of the Finance Secretary.

TRAINING SESSION FOR OFFICIALS

This circular follows a training session organized earlier by the Department of Personnel and Training (DoPT) for government officials, including those in the ranks of Joint Secretary and below. The training focused on using ChatGPT-4 Plus and was conducted from October 14-16, 2024. While it remains unclear whether other Ministries or departments received similar instructions, the DoE's circular specifically targets the use of AI tools in official devices.

WHY IT MATTERS?

The concerns surrounding DeepSeek have grown in recent weeks, with the app being scrutinized by regulators worldwide. Several countries have banned the use of DeepSeek's AI technology in government operations. Regulators in European nations, including Ireland, France, Belgium, and the Netherlands, have raised alarms about its data collection practices. Additionally, several federal agencies in the US have also instructed employees to avoid using the app, while many companies have asked their cybersecurity providers to block it, according to Bloomberg.

WHAT DEEPSEEK IS SAYING?

DeepSeek's privacy policy states, “We store the information we collect in secure servers located in the People’s Republic of China.” The primary concern with DeepSeek is that user data could potentially be shared with Chinese authorities, as China has laws requiring companies to comply with requests from local intelligence agencies to share data.

SO WHAT WILL INDIA DO?

In response to the concerns about DeepSeek, the Indian government is considering hosting the AI model on local servers. Union Minister of Electronics and Information Technology, Ashwini Vaishnaw, stated that data privacy issues related to DeepSeek could be addressed by hosting open-source models on Indian servers.

At the same time, India is making strides in AI development through the launch of the IndiaAI Mission. Sam Altman, CEO of OpenAI, recently visited India to push for OpenAI’s involvement in India’s national AI agenda. India has been playing catch-up in the AI race, especially after China’s DeepSeek gained attention for its low-cost AI model. In contrast, OpenAI has faced challenges with smaller rivals and legal disputes, such as a copyright lawsuit in India from publishers and media companies.

India is also a major market for OpenAI, being the second-largest after the US. During his visit, Altman mentioned that OpenAI had tripled its user base in India over the past year. Vaishnaw emphasized India’s ability to create cost-effective AI models, drawing comparisons to the country’s low-cost space missions.