The discovery of ChatGPT-generated text in a rain alert issued by the Pakistan Meteorological Department (PMD) has triggered a broader controversy, as new evidence shows that multiple government websites have also been using the AI tool. The incident is fueling concerns about the credibility of official communications and the lack of oversight in deploying artificial intelligence across ministries.
Intelligence analyst Zaki Khalid revealed that a specific search string could identify websites relying on ChatGPT-generated or edited content. Following the PMD case, the new findings pointed to several high-profile government portals and documents, including the China-Pakistan Economic Corridor (CPEC) website, the Higher Education Commission (HEC) policy on interim PhD placements, and the Finance Division’s Debt Sustainability Report for 2025–2027.
In some cases, URLs on government websites were discovered with the string “utm_source=chatgpt.com”. This UTM parameter, short for Urchin Tracking Module, is typically used for analytics to identify where traffic originates. Its presence on official .gov.pk pages suggests that either text or links generated by ChatGPT were copied and pasted directly, leaving behind the tracking code.
The growing number of government websites using ChatGPT has sparked debate about accountability and transparency. Critics warn that depending on generative AI for sensitive documents could result in errors, potential security risks, and an erosion of public trust. Supporters argue that AI can improve efficiency, but admit that clear guidelines are necessary to ensure accuracy.
Globally, the use of AI in public communication is on the rise, yet Pakistan’s experience highlights the risks of unregulated adoption. With more government websites linked to ChatGPT, pressure is mounting on authorities to establish a formal policy framework that governs AI use while protecting the credibility of official information.