Key Insights from GEM25 Panel at Harvard
At the recent GEM25 event hosted by Harvard’s Center for International Development, a panel of leading policy and technology experts from emerging economies gathered to discuss their visions for integrating Artificial Intelligence (AI) within their national strategies and societies. The panel featured Teresa Clarke, Chairman and CEO of Africa.com; Nanjira Sambuli, a nonresident scholar at the Carnegie Endowment for International Peace; Dr. Shikoh Gitau, CEO of Qhala; Umar Saif, Founder and CEO of aiSight.ai; and moderator Han Sheng Chia from the Center for Global Development.
Survey Highlights: Optimism and Concerns about AI in Africa
Teresa Clarke shared illuminating insights from a recent Africa.com survey, capturing the perspectives of respondents predominantly from Nigeria, Kenya, and South Africa. She noted that respondents were highly educated, with 78% holding college degrees. Notably, 78% expressed optimism about AI’s potential positive impact on Africa, significantly higher than comparable U.S. data.
Survey respondents highlighted education, healthcare, and agriculture as sectors poised to benefit most from AI adoption. Yet, there were clear concerns: job displacement, ethical challenges, and transparency issues topped the list. Interestingly, 55% of Africans surveyed were concerned about excessive government regulation of AI, compared to just 31% of Americans who shared this view.
Among use cases, Africans notably valued AI’s ability to enhance online product and service discovery and support agriculture through accessible, actionable data for farmers. Overall, 71% remained optimistic about AI’s transformative potential.
Respondents passionately contributed qualitative insights, emphasizing that localized, culturally-sensitive AI solutions are essential. One respondent articulated a common sentiment, stating, “Solutions should consider cultural, linguistic, and socioeconomic factors to effectively address the unique challenges faced by African communities.”
Electricity and Infrastructure: Critical Foundations for AI
A key theme echoed by Teresa Clarke was the critical importance of infrastructure—specifically electricity—to harness AI’s benefits. Clarke underscored that AI technology demands enormous power resources, noting that U.S. AI infrastructure alone consumes electricity comparable to Japan’s entire national usage.
“Without sufficient electricity and digital infrastructure, discussions about AI use cases become redundant,” Clarke emphasized. Investments in reliable, robust power supplies are paramount for emerging economies aiming to leverage AI technology meaningfully.
AI as a Strategic Partner, Not Just a Tool
A central shift identified by Clarke is how AI fundamentally transforms the relationship between technology and business or government leaders. Unlike previous technologies handled by IT departments, AI demands a strategic partnership at every decision-making level, closely resembling a “council of elders,” offering insights based on extensive data analysis and pattern recognition.
Clarke provided vivid examples from African telecommunications companies, highlighting how each entity strategically applies AI based on their objectives—be it cost reduction or service enhancement. This underscores AI’s versatile role, tailored uniquely to different organizational priorities.
Balancing Data with Human Insight
Panelist Nanjira Sambuli highlighted the continued importance of human context and cultural insight alongside AI technology. For instance, Kenyan fintech solutions initially aimed to limit nighttime lending to curb gambling, only to discover through human insight that market women also relied on nighttime loans for early business activities. Such examples illustrate how human intelligence remains crucial to interpreting AI-generated insights appropriately.
Sambuli also emphasized the limitations of AI in replicating human trust dynamics, referencing the success of traditional savings groups in marginalized communities, which AI-driven fintech solutions have struggled to match.
Representation and Data Equity
The panel addressed concerns about bias in AI systems, stressing the importance of diverse representation in AI development. Clarke cited Joy Buolamwini’s work on facial recognition biases to illustrate how lack of representation among coders can result in AI systems that fail to recognize diverse faces accurately. Ensuring diverse participation in coding processes is essential to mitigate such biases and create AI systems equitable and relevant across cultures.
Localization of Data Assets
The panel further underscored the significance of localizing data assets, likening data to valuable commodities such as gold. Clarke explained that data centers’ physical proximity significantly impacts data access speeds, emphasizing the necessity for Africa and emerging economies to develop localized data infrastructure to ensure competitive parity globally.
Ultimately, the GEM25 panel highlighted a future where AI holds immense promise for emerging economies—provided foundational issues around infrastructure, representation, and localized data ownership are effectively addressed.
Watch the full panel discussion below.
