Amid the Publican AI controversy, claims by the Ghana Revenue Authority (GRA) that a newly deployed AI system has boosted customs revenue by nearly 50% within just two weeks are coming under growing scrutiny.
According to analysts, the numbers churned by GRA as a result of Publican AI may be too good and too early to celebrate.
Data and policy analyst Alfred Appiah is urging caution, arguing that without deeper technical validation and proper auditing, the reported surge in revenue cannot yet be taken as proof of success.
At the centre of the debate is the AI-powered tool introduced at the ports to enhance revenue mobilisation. The GRA has pointed to sharp gains to the tune of $3 million in collections shortly after its rollout.
However, Alfred Appiah says the headline figure raises more questions than answers.
Is Two Weeks Enough?
For the analyst, one of the key concerns is timing. He explains that customs revenue does not always follow a steady pattern. It can fluctuate based on shipment volumes, import cycles, and seasonal trade activity.
A spike within a two-week window, Appiah argues, may reflect temporary factors rather than a structural improvement driven by the AI system. He argues that without understanding the baseline and seasonal trends, it’s difficult to attribute the increase solely to the new system.
“There are already claims of strong results. The GRA reports that customs revenue increased by nearly 50% within two weeks of implementation. This raises questions about whether customs revenue is consistent every two weeks and whether seasonality plays a role,” he noted.
Overcharging or Efficiency?
There are also concerns from industry players about possible overvaluation or overcharging at the ports, an issue that could artificially inflate revenue figures.
For him, if importers are being assessed higher duties than expected, revenue may rise in the short term, but at the cost of trust and compliance in the long run.
This, Appiah notes, makes it critical to independently audit the system’s assessments to determine whether the gains are coming from improved efficiency or questionable valuations.
He said, “With industry players already raising concerns about overcharging, it’s unclear whether this revenue increase can be considered a reliable measure of success without a proper audit of the AI system’s assessments.”
Who Is Behind Publican AI?
Beyond the numbers, questions are also being raised about the entities involved in selling the AI solution to Ghana.
According to Alfred Appiah, Publican is reportedly developed by a startup known as Ultra Information Solutions. However, the government’s contract is said to be with another entity, Truedare Investment Services.
This has sparked concerns about transparency, with some observers asking whether intermediaries are playing a role in the deal and what value they bring.
Do We Need Another System?
Perhaps the most fundamental question is whether Publican is even necessary. Ghana already operates the Integrated Customs Management System (ICUMS), managed by Ghana Link Network Services, which was designed to handle customs valuation, risk management, and revenue assurance.
Alfred Appiah questions whether ICUMS is incapable of delivering similar results or whether enhancements, including AI features, could have been integrated into the existing system instead of introducing a new layer.
This scepticism is also shaped by recent history. Some critics recall similar claims of dramatic revenue improvements tied to technology interventions in the past, claims that later faced scrutiny.
This has made analysts wary of celebrating early results without robust, verifiable evidence.
