UK Home Office faces probe threat over secret ChatGPT use

The UK’s Home Office could be investigated by a government watchdog over its use of OpenAI’s ChatGPT technology to process asylum claims. Wired reports that concerns focus on the accuracy of the AI tool and the lack of transparency around how it is being used in these sensitive decisions.

Why it matters

When a government department uses AI to help decide asylum claims, it sets a precedent for how automated tools are trusted with high-stakes decisions. Wired says the UK’s Independent Commissioner’s Office is considering a formal investigation, which could force the Home Office to reveal how it uses ChatGPT and what safeguards are in place.

The core issue is trust. As more organisations rely on AI tools, the pressure to show how decisions are made, and to prove their accuracy, grows. For the Home Office, public scrutiny is especially intense because asylum decisions can be life-changing, and mistakes carry heavy consequences. According to Wired, critics are concerned that relying on a commercial AI model introduces risks of bias, errors, and a lack of accountability.

Watchdog focuses on secrecy and accuracy

The main complaint is that the Home Office has not explained how ChatGPT is being used in asylum processing. Wired reports that the government has not published details about what data is fed into the system, what checks are in place, or how human oversight works. This secrecy has triggered calls for an investigation from advocacy groups and lawmakers.

Wired notes that the UK's data regulator, the Information Commissioner’s Office, is now weighing whether to launch a full probe. If that happens, the Home Office could be compelled to disclose technical documentation, audit logs, and the results of any internal reviews. These steps could reveal whether the use of ChatGPT aligns with UK data protection and fairness laws, or if changes are needed.

For businesses, this process is a warning sign. If even a government department faces scrutiny for using off-the-shelf AI models, private companies using tools like ChatGPT for customer service, hiring, or compliance may face similar questions from regulators, customers, or staff. Recent coverage shows that AI adoption without transparency can expose firms to operational and reputational risk.

What this means if you run a business

Any company using third-party AI tools, especially those handling sensitive data or decisions, should expect rising pressure to open the black box. If regulators investigate the Home Office, that will likely set new expectations for disclosure and oversight in both the public and private sector.

For business owners, this means documenting how AI tools are deployed, keeping records of inputs and outputs, and being ready to explain how decisions are made. Relying on vendors like OpenAI does not remove your responsibility for the outcome. If your company uses ChatGPT or similar models in operations that affect people’s rights, privacy, or finances, you should be prepared for questions from regulators, clients, or advocacy groups.

This case shows that secrecy around AI use is itself a risk. Companies that proactively disclose how they use AI, and what guardrails are in place, will be better positioned if scrutiny comes. As AI becomes part of more business processes, transparency and accountability are not just compliance issues, they are essential for trust.