Published 14:11 IST, February 5th 2025
Finance Ministry Warns Employees Against Use of ChatGPT, DeepSeek R1
Countries like Australia and Italy have placed similar restrictions on the use of DeepSeek, citing data security risks.

The Ministry of Finance has asked its employees to avoid using AI tools including ChatGPT and DeepSeek for official purposes, citing risks posed to the confidentiality of government documents and data, an internal department advisory showed.
"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," said the advisory by the finance ministry dated January 29.
Three finance ministry officials told Reuters that the note was genuine and the note was issued internally this week.
Countries like Australia and Italy have placed similar restrictions on the use of DeepSeek, citing data security risks. Reports of the advisory surfaced on social media on Tuesday, ahead of a scheduled visit to India by OpenAI chief Sam Altman on Wednesday, when he is also due to meet the IT minister.
The advisory comes days after IT Minister Ashwini Vaishnaw said DeepSeek R1, built by a Chinese company called DeepSeek, will soon be hosted on Indian servers to address data privacy concerns.
“The good thing is that Deepseek is an open-source model, and we are very soon going to host DeepSeek on Indian servers, the way we have hosted Llama (large language model/generative AI model developed by Meta AI) on India servers,” said Vaishnaw.
He added that every app or system that needs to be evaluated for its data privacy policy and assessed on data security parameters will be tested, based on which a decision will be taken. “Everything that is open source can be taken and hosted on our server so data privacy parameters can be addressed, and that we are going to do soon," the minister said.
According to the minister, the specifics such as the number of servers required to host the DeepSeek V3 foundational model “have been worked out.”
– Written with inputs from Reuters
Updated 14:11 IST, February 5th 2025