Kenya’s professional services firms are keeping off artificial intelligence (AI) for fear of crossing ethical and legal lines that could cost them their licences.
Dipesh Shah, the Chief Executive Officer of Grant Thornton Kenya, said that accountancy and audit firms are especially wary of using open AI platforms such as ChatGPT due to the risk of breaching client confidentiality and data privacy laws.
“We sit on a lot of client confidentiality matters and data. You can’t just upload it to ChatGPT and get a report — you’re not allowed,” Mr Shah told the Business Daily in an interview.
“In fact, we’re holding back on AI integration because of that. We need the right solution to ensure we don’t infringe on this privacy,” he added.
According to Mr Shah, the key concern is that ChatGPT is an open-model AI developed by OpenAI — meaning user details may be processed and stored on systems outside the firm’s control.
“Because this is open AI, the information will somehow be made public,” he cautioned.
OpenAI is a San Francisco-based AI research and technology company founded in December 2015 by billionaire Elon Musk, Sam Altman, Greg Brockman, Ilya Sutskever, Wojciech Zaremba, and John Schulman.
Its flagship product, ChatGPT, can generate human-like text, summarise documents, and answer queries in natural language. OpenAI’s mission is to ensure artificial general intelligence benefits all of humanity.
A recent report by global internet researchers Datareportal found that Kenyans are among the world’s most enthusiastic ChatGPT users.
The study showed that 42.1 percent of Kenyans aged over 18 use ChatGPT for daily tasks — the highest penetration rate globally.
Yet, while the public embraces the tool, professionals such as accountants, auditors, and lawyers face strict restrictions. The Accountants (Standards of Professional Practice and Ethical Conduct) Regulations explicitly require practitioners to protect client data:
“An accountant shall respect the confidentiality of information acquired as a result of professional and business relationships and shall not disclose any such information to third parties without proper and specific authority unless there is a legal or professional right or duty to disclose.”
Breaching this clause can attract stiff penalties, including fines or suspension of a practising certificate.
This is why, Shah explained, Grant Thornton and other firms are exploring closed-model AI systems — platforms where the underlying model and training data remain exclusive, and client inputs stay within a secure, private environment.
Closed AI systems allow firms to harness automation and advanced analytics without exposing sensitive data to public networks. By contrast, open AI models such as ChatGPT, Google’s Gemini, or Claude, owned by former OpenAI employees, operate via external servers and shared infrastructure, making them unsuitable for handling confidential information in regulated industries.
Despite the caution, Shah says the profession cannot ignore technology. He said Grant Thornton is looking at creating secure, in-house databases to support AI-powered tools in audit, tax, and advisory work.
“Our profession is very restrictive. Ethics don’t allow us to infringe client confidentiality,” he said, “but we also have to embrace technology.”
As Kenya’s AI adoption accelerates across banking, manufacturing, and retail, professional services firms are working to find that sweet spot — leveraging innovation without compromising ethics or violating the law.
