TAGG criticises GRA over Publican AI rollout, expresses disappointment in Commissioner-General
13th April 2026
The General Secretary of the Traders Advocacy Group Ghana (TAGG), Nana Poku, has criticised the Commissioner-General of the Ghana Revenue Authority (GRA), Anthony Sarpong, over the implementation of the Publican Artificial Intelligence (AI) system at Ghana’s ports.
Speaking on Adom FM’s Dwaso Nsem morning show, he questioned the technical foundation of the system, particularly its reliance on third-party data sources.
“I am very disappointed in the GRA boss. With the level of knowledge I expected from him, I didn’t expect this approach,” he said.
According to him, concerns raised by traders suggest that the system is built on an external platform that may not adequately reflect Ghana’s trade environment.
“We came to explain that the software is a third-party system and cannot fully serve the country’s needs if it is not properly structured,” he noted.
Nana Poku stressed that data integrity is key to the effectiveness of any artificial intelligence system.
“AI without proper data cannot speak. If the data is wrong, the output will also be wrong,” he said.
He further alleged that the system operates through an intermediary platform linked to a Canadian company, insisting that Ghana should have greater control over its own data infrastructure.
“We are saying Ghana should go directly to the original source instead of relying on a middle platform. That is our concern,” he added.
The TAGG General Secretary emphasised that traders are not opposed to reforms but are calling for transparency, fairness, and accountability in the system’s rollout.
“What we want is accountability and fairness. Government must move forward with a system that allows traders to work freely and confidently,” he said.
Meanwhile, GRA Commissioner-General Anthony Sarpong has defended the system, rejecting claims that it has slowed down operations at the ports. He maintains that the Publican AI system has instead improved efficiency compared to the previous manual processes.