An international law firm has restricted access to several artificial intelligence (AI) tools after noticing a "significant increase in usage" by its staff.
In an email seen by the BBC, a senior director at Hill Dickinson, which employs over a thousand people in the UK, cautioned staff about using AI tools.
The firm stated that much of the usage did not comply with its AI policy. Moving forward, staff will need to request access to these tools.
A spokesperson from the Information Commissioner's Office, the UK's data watchdog, told BBC News that firms should not discourage the use of AI at work.
The spokesperson added: "With AI providing many ways to work more efficiently and effectively, the solution cannot be for organizations to ban AI use and push staff to use it secretly."
"Instead, companies should provide their staff with AI tools that align with their organizational policies and data protection obligations."
In the email, Hill Dickinson's chief technology officer reported that the law firm detected over 32,000 visits to the popular chatbot ChatGPT during a seven-day period in January and February.
During the same period, there were also more than 3,000 visits to the Chinese AI service DeepSeek, which was recently banned from Australian government devices due to security concerns.
The email also noted nearly 50,000 visits to Grammarly, the writing assistance tool.
However, it is unclear how many times staff visited ChatGPT, DeepSeek, or Grammarly, or how many staff members visited repeatedly, as multiple visits could have been made by a single user each time they accessed the websites.
The email to Hill Dickinson employees stated: "We have been monitoring the use of AI tools, especially publicly available generative AI solutions, and have noticed a significant increase in their use and the uploading of files to these tools."
Hill Dickinson, which has offices in various parts of England and abroad, later told BBC News: "Like many law firms, we aim to positively embrace the use of AI tools to enhance our capabilities while always ensuring their safe and proper use by our staff and for our clients."
The firm added that its AI policy, which includes guidelines prohibiting the upload of client information and requiring staff to verify the accuracy of responses from large language models, will ensure that usage remains "safe, secure, and effective."
The firm is now granting access to AI tools only through a request process. It is understood that some requests have already been received and approved.
A spokesperson from the Solicitors Regulation Authority, which oversees solicitors in England and Wales, told BBC News: "Despite the growing interest in new technology, there is still a lack of digital skills across all sectors in the UK. This could pose a risk for firms and consumers if legal practitioners do not fully understand the new technology being implemented."
A survey of 500 UK solicitors by legal software provider Clio in September found that 62% expected AI usage to increase over the next 12 months. It revealed that law firms across the UK were using the technology for tasks like drafting documents, reviewing or analyzing contracts, and conducting legal research.
A spokesperson from the Department for Science, Innovation and Technology described AI as a "technological leap" that will "free workers from repetitive tasks and unlock more rewarding opportunities." They told BBC News: "We are committed to introducing legislation that allows us to safely harness AI's enormous benefits. We are engaging widely and will launch a public consultation soon to ensure our approach effectively addresses this rapidly evolving technology."