Artificial Intelligence will be a defining technology in Safaricom’s push to become a technology company by 2030, but the operator is still at an early stage and spent much of the first half of its financial year building the required foundations, Chief Technology Officer James Maitai said on the sidelines of the company’s half-year results announcement.
Safaricom AI groundwork: unified data platform and skills
Maitai said Safaricom is “just scratching the surface” of AI adoption. The company has built a unified data platform (UDP) that brings its data together in one place and runs in a modern data centre. “As you well know, AI runs on data,” he said, adding that UDP is the first step toward wider AI use.
He identified three prerequisites for successful AI: skilled people, strong data governance and clear use cases. Safaricom has trained engineers, data scientists and data engineers and established what Maitai described as a “very strong data governance framework.” The operator has also hired more data scientists and retrained existing staff with technical backgrounds.
Operational use cases and customer-facing AI
Within technology, Safaricom is using AI for intelligent operations (AIOps) to manage its large and growing network and transaction volumes. Maitai noted the scale: M-PESA processed 27.4 billion transactions in the six months to September 2026, up 25.2% year on year. The company is also applying AI to “smart capex” for more precise capital spending decisions — “You put every shilling where it’s supposed to be,” he said.
Outside engineering, Safaricom already uses AI for customer value management and targeted offers and has begun deploying AI in customer care. The first customer-facing deployment is Zuri, the chatbot available through the My Safaricom app. Maitai said this is only the start for call centres: the company is building AI tools to increase call handling capacity, improve customer experience and enhance agent experience.
Safaricom has created a dedicated innovation team within its AI unit to explore emerging technologies, rapidly prototype and test models, and shorten product time-to-market. A key focus is reducing friction in customer journeys by identifying where customers struggle and cutting the number of steps needed to complete tasks.
Hybrid model approach and partnerships
On model strategy, Maitai said Safaricom is taking a hybrid approach. “It will be a mix of both. So there are some models we take and adapt for internal use and for serving our customers,” he said, while also partnering with global companies in areas where those firms have clear advantages.
Connectivity investments and capacity
Alongside AI, Safaricom has heavily invested in network capacity this year. Maitai described connectivity as “a universal human right” and said customers are increasingly sophisticated and data-hungry. In the first half of the year the company built more than 500 new sites, including more than 200 5G sites, connected more than 120,000 new homes to fibre, expanded capacity on almost 2,000 existing sites and upgraded its transport backbone and caching capability.
Safaricom front-loaded these investments into the first half so commercial teams could sell on new capacity while the technology team uses the second half to optimise the network and plan for the following year. The operator became the first Kenyan company to pass one million fixed data subscriptions earlier this year, holding 36.1% of the market, according to the Communications Authority of Kenya.
Maitai emphasised that technical investment must be matched by investment in people and processes. The unified data platform also enables engineers to learn by experimenting, complementing classroom training: “So it’s a hybrid of both,” he said. “Without investing in your people, without investing in data governance processes, it doesn’t make sense to invest in just the technical capability in terms of the equipment. So people are key.”

