Physical Address
60 Ekwema Cres, Layout 460281, Imo
Physical Address
60 Ekwema Cres, Layout 460281, Imo
African AI startups have raised about $1.3 billion in disclosed funding over the past decade, but most are still struggling to make money. Only 15 percent are profitable, according to a new report from German development agency GIZ and data firm Briter.
The report, titled “AI Made in Africa: Voices from AI businesses across Africa”, mapped 311 AI companies across the continent. Its main finding is that building AI in Africa is expensive, while selling it is hard.
The biggest problem is cost. For 65 percent of the companies surveyed, cloud computing, GPUs and AI infrastructure are their largest development expense, and these bills are paid in US dollars. Some startups serving thousands of users a day spend $20,000 to $50,000 a month on third-party AI APIs alone.
At the same time, more than half of the companies charge customers less than $2,500 a year, or about $208 a month. That gap between dollar costs and local prices squeezes margins from both sides.
Currency swings make it worse. The report notes that Nigeria’s 2023 naira devaluation reportedly raised local cloud bills by 224 percent for some users.
Part of the problem is infrastructure. Africa is home to about 18 percent of the world’s population but holds less than 1 percent of global data centre capacity, so most AI computation for African products runs on servers abroad.
Funding is also concentrated. More than 75 percent of the capital raised went to companies founded before 2020. The median funding round was just $225,000, and about one-third of the companies are fully bootstrapped.
Read Next: Nigerian Banks Face January 2027 Data Localisation Deadline as NITDA, CBN Push Sovereign Cloud
Selling AI is another challenge. Some 65 percent of companies say customer trust and AI literacy are their biggest barrier to growth, even ahead of development costs. Around 44 percent sell to governments, but those deals can take 12 to 18 months to close.
Talent is scarce too. Africa has roughly 1 percent of the global AI talent pool, and 51 percent of founders say senior AI research roles are the hardest to fill. Nearly half have built their own in-house training programmes.
The report arrives as funding for African AI is falling. Pan-African investor Enza Capital, which this month announced a $21 million first close for a new AI Africa Fund, cited data showing that disclosed investment in African AI companies fell by more than 90 percent between 2025 and the first half of 2026.
The authors call for more patient capital from investors who understand African operating conditions. They also urge governments to lower the cost of key inputs like compute and connectivity, rather than simply celebrating the sector’s growth.
For founders, the lesson is clear: in African AI, cost control and pricing strategy may matter as much as the model itself.
Source: GIZ and Briter, “AI Made in Africa”, as reported by WeeTracker.
Was this information useful? Drop a nice comment below. You can also check out other useful contents by following us on X/Twitter @siliconafritech, Instagram @Siliconafricatech, or Facebook @SiliconAfrica.