Ghana is on the verge of a legislative tangle that could cripple its technology ecosystem before it even finds its feet, says Sitsofe Mensah, a technology policy analyst with the IMANI Centre for Science, Technology and Innovation Policy.
Sitsofe argues that the country’s current approach to AI regulation is “importing confusion, not clarity.”
According to the policy analyst, Ghana is attempting to align its upcoming Emerging Technologies Bill (ETB) with the European Union’s celebrated AI Act, which is a global regulatory gold standard.
For him, the intention is sound since it aims to ensure that Ghana’s digital laws meet the highest global standards. But the execution, Mensah cautions, is dangerously flawed.
Three Bills, One Problem
The article by the technology policy analyst notes that currently, three major technology-related bills are being drafted or considered.
These are the Emerging Technologies Bill, the Cybersecurity (Amendment) Bill, and the Data Protection (Amendment) Bill.
Each bill, on its own, serves a useful purpose. The ETB aims to govern the ethical use of emerging technologies like Artificial Intelligence. The Cybersecurity amendment gives the Cyber Security Authority power to certify the safety of AI systems. The Data Protection amendment updates the country’s data privacy rules.
However, according to IMANI’s analysis, the real danger lies in how these laws interact or fail to. Sitsofe Mensah explains that the three overlapping laws do not speak to each other. Each agency is being given power over AI from a different angle, setting up a three-way jurisdictional war before a single law is even passed.
He warns that this duplication of authority could stall innovation, discourage investment, and leave startups tangled in legal uncertainty.
Missing Link: The “Bias Safe Harbor”
The expert says perhaps the most technical, but critical gap lies in what he calls the “Bias Paradox.” Both the ETB and Ghana’s National AI Strategy insist that AI systems must be “bias-free and ethical.”
But under the current legal framework, testing for bias in an algorithm could itself be illegal.
For instance, to check whether a loan approval system discriminates against women or ethnic minorities, developers need to process sensitive personal data like gender or ethnicity. Under Ghana’s existing Data Protection Act (Act 843), doing so is a legal risk.
The EU solved this exact problem through Article 10(5) of its AI Act, which is a narrow legal exception that allows developers to process sensitive data under strict safeguards, but only for detecting and correcting bias.
Ghana’s draft, Mensah says, copied the rule requiring bias-free AI but forgot to include the very clause that makes it legally possible. Without this fix, any effort to make AI “fair” could inadvertently break the law.
What Must Be Done: IMANI’s Practical Roadmap
IMANI’s recommendations are clear and pragmatic. Mensah calls on the Ministry of Communication, Digital Technology, and Innovations (MoCDTI), Cabinet, and Parliament to intervene now before the bills are finalized.
- Define a Clear Chain of Command:
The government must clearly state which agency leads AI regulation and which plays supporting roles. A simple “Statement of Interoperability” could prevent turf wars between agencies.
- Insert the “Bias Safe Harbor”:
The ETB must include a Ghanaian version of the EU’s Article 10(5) to allow developers to process sensitive data safely for the limited purpose of auditing AI systems for bias.
- Protect Startups from Overregulation:
The ETB should be aligned with the Innovation and Start-Up Bill, ensuring that compliance requirements are proportional. Startups should not be crushed under costly paperwork before they even generate revenue.
- Cabinet and Parliament Oversight:
Before approving any of the bills, Cabinet and Parliament should insist that all three, ETB, Cybersecurity, and Data Protection, are coherent and interoperable.
The Bottomline
IMANI’s Sitsofe Mensah says Ghana must not legislate its way into confusion.
For him, passing three conflicting laws at the same time is not regulation, but rather chaos. He cautions that if the government truly wants to harness AI for national development, then the laws must be as smart, agile, and coordinated as the technology itself.
As it stands now, Ghana stands at a critical crossroads; one path leads to innovation and inclusive growth, the other to bureaucratic paralysis.
