Understanding the Risks of AI in Defense

The rapid advancements in artificial intelligence (AI) have brought tools like ChatGPT and DeepSeek AI into the spotlight. Their natural language processing and content generation capabilities make them valuable across various sectors. However, concerns over data privacy, national security, and operational secrecy remain paramount for defence organisations.

No Official Ban, But Caution is Advised

Currently, there is no explicit restriction on Indian Army officers’ use of AI platforms like ChatGPT or DeepSeek. However, strong concerns exist regarding operational secrecy. Any data entered into these AI tools could be stored, processed, or analyzed on servers outside India’s control, raising the risk of data leaks or unauthorized access.

Indian Army officers hold sensitive positions where confidentiality is critical. While AI chatbots may assist in summarizing reports or drafting documents, the risk of exposing classified or strategic information far outweighs any convenience these tools offer.

Ministry of Finance’s Directive on AI Use

The Ministry of Finance recently prohibited employees from using AI tools on office devices, including ChatGPT and DeepSeek. A directive issued on January 29 stated:

“It has been determined that AI tools and AI apps (such as ChatGPT, DeepSeek, etc.) in office computers and devices pose risks for the confidentiality of government data and documents.”

Given that the Ministry has already taken this stance, it is likely that other government departments—including the defence sector—will adopt similar cautionary measures to mitigate risks.

DeepSeek: A Rising AI Competitor with Security Concerns

While OpenAI’s ChatGPT remains a dominant force in AI, DeepSeek has gained traction due to its cost-effective model. Reports indicate that DeepSeek R1 operates with significantly lower computing costs than established AI systems. However, concerns over privacy and data security have led countries like Australia and Italy to restrict the use of Chinese-developed DeepSeek in government operations.

In an era where data is considered the “new oil,” any potential vulnerability in AI systems is a significant concern, especially for military and government institutions handling sensitive information.

The Global AI Landscape and Emerging Regulations

AI regulations are still evolving, and they are struggling to keep pace with rapid technological advancements. Even in the United States, where many leading AI models originate, there is increasing scrutiny over data privacy, ethical AI usage, and security frameworks.

Sam Altman, CEO of OpenAI, recently visited India to discuss AI adoption with government officials. While India is exploring AI-driven innovation, the government remains cautious about potential risks, particularly in classified data sectors.

Key Risks for the Indian Army

• Operational Secrecy: Any breach of military strategies, deployment plans, or logistical details could pose a direct threat to national security.
• Lack of Clear Guidelines: Although no explicit ban exists, cautionary precedents set by other government departments suggest that restrictions may follow.
• Risk of Data Leaks: AI tools process vast amounts of data, and inadvertent sharing of sensitive details could expose critical intelligence.
• Intelligence Gathering by Adversaries: Even seemingly harmless data—such as supply chain information or routine logs—can be pieced together by hostile actors to gain strategic insights.

Future of AI in Defense

While AI tools currently pose security risks, the future may see the development of in-house, defence-specific AI solutions. These systems would be designed to operate within government-controlled infrastructure, ensuring compliance with national security protocols. Until then, reliance on external AI platforms remains a significant risk.

Conclusion: Exercise Caution Until Clear Policies Emerge

While there is no outright ban on AI tools like ChatGPT and DeepSeek for Indian Army officers, the risks associated with data security and operational secrecy strongly discourage their use for official tasks. The Ministry of Finance’s directive underscores the government’s stance on prioritizing data confidentiality over convenience.

For now, the safest approach for Army personnel is to avoid AI tools, particularly on official devices or for sensitive tasks, until the Indian government establishes robust security frameworks and usage guidelines.