Artificial Intelligence Needs a Defined Job
To harness artificial intelligence effectively, Liberia must first define its specific job, starting with small, accountable tasks that build capability rather than chasing transformative but ill-defined ambitions.

Editorial analysis. This piece reflects Executive Tea Chronicle's interpretation and editorial mission, not an official government position.
Analysis. Drafted from cited World Bank sources for editor review; Liberia proposals are the magazine's analysis, not source findings.
Source scope: the cited sources describe historical project designs and financing announcements (2021–2024), not verified current implementation outcomes. No new Liberia-specific factual evidence was checked for this draft; Liberia scenarios are hypothetical.
The global conversation around artificial intelligence often focuses on radical transformation. For developing economies, this can present an alluring prospect: the chance to bypass legacy systems, improve administrative processes, and unlock new avenues for growth. However, technology without a clear purpose can become an expensive distraction. For a country like Liberia, considering investment in complex AI systems requires a fundamental question: what specific, well-defined job would the technology be hired to do? The answer should point not to a vague aspiration for “efficiency,” but to a concrete operational task.
An analysis of digital strategies in the region suggests that sustainable progress often comes from building a solid foundation first. While not an AI-specific initiative, Ghana’s digital agenda provides a useful case study in sequencing. Its focus has been on the groundwork that makes advanced tools useful. A key lesson from such an approach is not about replicating a specific program, but embracing the principle of building capability before expanding complexity. The adoption of AI could therefore be a deliberate, incremental process, beginning with tasks where benefits are clear and risks are manageable.
What the comparison shows
Ghana’s strategy, supported by international partners, focuses on the fundamentals of a digital economy. According to a World Bank press release of 28 April 2022, the institution approved a $200 million project to accelerate the country’s digital transformation. The stated goals of this Ghana Digital Acceleration Project are not centered on a single, advanced technology like AI, but on creating an enabling environment. The project design aims to increase access to mobile internet and broadband services for six million people, particularly in underserved rural areas, by encouraging private investment.
The project builds on previous efforts, like the e-Transform Ghana project, to enhance the delivery of digital public services. Its objectives, as described by the World Bank, include streamlining government processes and strengthening the local digital innovation ecosystem by supporting startups and promoting advanced digital skills. A key component of the 2022 project financing was facilitating smallholder engagement in data-driven digital agriculture. This comprehensive approach underscores a critical insight: advanced digital tools may be most effective when built upon a base of widespread connectivity, digital literacy, and modernized public services. The project was designed to address bottlenecks identified in a 2020 diagnostic of Ghana's digital economy, which the World Bank noted had grown an average of 19 percent annually between 2014 and 2020.
Analysis: applying the lesson in Liberia
For Liberia, a potential lesson is to begin with a narrow, clearly defined scope. Rather than pursuing a national AI strategy, a single government ministry or public corporation could hypothetically pilot a tool for a specific, low-risk internal function. Such an approach could help build institutional capacity and reveal practical challenges without committing significant resources to a concept yet unproven in the local context. Functions like regulatory compliance or aspects of public procurement could serve as potential starting points.
An agency could hypothetically deploy an AI-powered tool for one of two initial jobs:
- Document Search and Retrieval: The task of searching for specific clauses in laws, past contracts, or policy manuals can be time-consuming. A simple AI tool, trained exclusively on an agency’s own approved documents, could perform this task quickly. In a hypothetical public works scenario, for instance, an engineer could use such a tool to find all approved specifications for a particular construction material, saving research time. The final decision would remain with the engineer, but the information retrieval process could be accelerated.
- Assisted Drafting: Many organizations produce repetitive documents like permits, official correspondence, or preliminary reports. An AI assistant could be designed to generate a first draft based on a template and user-provided data points. In a hypothetical example, if a business applied for a standard permit online, an AI tool could pre-populate the permit form using the submitted information, flagging any missing fields for a human officer to review and approve. The officer's role could then shift from manual data entry to verification and oversight.
In both cases, the AI is positioned as an assistant, not an autonomous decision-maker. The system would work with a closed set of approved information, and a human being would be explicitly responsible for the final output.
Tradeoffs and implementation risks
Adopting even simple AI tools requires a clear-eyed assessment of risk. A first step for any pilot project would be to define acceptable performance and error rates. For example, if an AI tool for summarizing reports is correct an illustrative 95% of the time, decision-makers would need to determine if that is sufficient for its intended purpose. The answer would depend entirely on the consequences of the remaining illustrative 5% of cases where it is incorrect.
Data privacy and security are paramount. Any system, especially one handling government information, must have robust safeguards to protect sensitive commercial or citizen data. Responsibility must be unambiguous: the head of the agency remains accountable for all decisions and outputs, regardless of what tools their staff uses. The phrase “the algorithm made a mistake” cannot be an acceptable excuse for mission failure or public harm.
This is why high-consequence uses—such as determining citizen eligibility for social benefits, assessing tax liability, or flagging individuals for law enforcement scrutiny—demand a much higher standard of scrutiny. Such systems would require rigorous, independent pre-deployment assessment and, crucially, a clear, accessible, and effective process for citizens to appeal and correct machine-driven errors.
What global competence looks like here
In the 21st century, technological competence is not simply the ability to procure and switch on the latest gadget. It is the wisdom to know where an algorithm helps and where human judgment is irreplaceable. It means understanding a tool’s limitations as well as its strengths. A truly competent organization is one that can run a disciplined pilot, measure its impact against a clear baseline, and make an evidence-based decision to either scale, modify, or terminate the project.
This approach—starting small, defining the job, insisting on human accountability, and measuring results—is the hallmark of a serious institution. It demonstrates a commitment to operational excellence over technological theater. It proves that the organization is focused on solving a real-world problem for the people it serves, not just on appearing modern.
Bottom line
Artificial intelligence is not a magic wand; it is a powerful but specialized tool that functions best with a specific job description. For Liberia, a potential path forward could involve learning from the foundational logic evident in Ghana’s digital acceleration plans. This analysis suggests that a prudent approach would focus on building basic digital infrastructure and capacity. If AI is introduced, it could be through controlled pilots with accountable owners, tasked with solving discrete, low-risk problems. Proving the value of AI in assisting human experts with tangible work could build the experience and trust needed to tackle more complex challenges in the future.
Sources
- World Bank Provides $200 Million to Accelerate Ghana’s Digital Transformation Agenda for Better Jobs — World Bank, published 2022-04-28
Related reading

Design Digital Services Around a Completed Transaction
Inspired by the design of Rwanda's digital agenda, Liberia's public agencies should build online services that deliver a complete result, not just collect an application, to truly modernize and respect citizens' time.

Shared Data Standards Can Reduce Administrative Friction
Kenya’s plan for a data-driven public sector shows how shared technical standards could reduce the administrative friction that slows Liberia’s economy, starting with a pilot project between key agencies.

Digital Inclusion Requires Assisted Access
For Liberia's digital services to succeed, the country must learn from international examples and build assisted access points to bridge the gap for citizens with limited literacy, devices, or connectivity.