Every Software Tool Your Business Uses Has Probably Added AI In The Last 12 Months & What This Means Going Forward
Akili AI executives Simon Bransfield-Garth and Shikoli Makatiani detail how autonomous Agentic AI is replacing reactive generative models in Kenyan finance.
The era of rudimentary artificial intelligence acting merely as a conversational parlor trick is over. According to Simon Bransfield-Garth, CEO of Akili AI, and Chief Technology Officer Shikoli Makatiani, the African enterprise sector is rapidly shifting from generative models to autonomous Agentic AI. The transition fundamentally redefines how Kenyan banks, microfinance institutions, and cooperatives handle risk, compliance, and capital allocation.
Speaking on the architectural divide between the two technologies, the Akili AI executives outlined a brutal reality for corporate boards: organizations failing to govern their internal AI usage are actively compounding their risk registers, while those harnessing agentic systems are securing massive, asymmetrical market advantages.
Defining the Divide: Generative vs. Agentic
The core difference between the two systems lies in autonomy and execution. Generative AI, which dominates current headlines, is fundamentally reactive. It synthesizes data, writes emails, and summarizes reports only when prompted by a human operator. It is a sophisticated oracle, but it cannot take independent action.
Agentic AI, conversely, is proactive and goal-oriented. As Bransfield-Garth explained, an agentic system is granted specific authority within a digital workflow. In a Kenyan financial context, a generative AI might summarize a loan applicant’s financial history. An Agentic AI, however, will independently pull data from the Credit Reference Bureau (CRB), analyze M-Pesa transaction volumes, verify KRA PIN compliance, structure the loan terms, and execute the approval—all within seconds, without human intervention.
- Generative AI: Reactive, conversational, requires constant human prompting.
- Agentic AI: Autonomous, executes multi-step workflows, directly interacts with external APIs.
- Financial Impact: Reduces loan origination times from days to milliseconds.
The Menace of Shadow AI
A critical vulnerability exposed during the discussion is the explosion of “shadow AI.” Makatiani highlighted that many Kenyan executives falsely believe their organizations have not yet adopted AI because they haven’t officially procured enterprise software. In reality, employees across customer service, HR, and marketing are independently feeding sensitive corporate data into public, unregulated models to accelerate their daily tasks.
This shadow deployment creates catastrophic governance and data sovereignty risks. If a loan officer uploads proprietary financial algorithms or unanonymized client data into a public generative model, the institution instantly breaches stringent Central Bank of Kenya (CBK) data privacy regulations. Akili AI advocates for immediate internal audits and the deployment of hardened, ring-fenced enterprise models that keep corporate data securely within institutional walls.
Deploying Autonomous Agents in Kenyan Banking
The practical application of Agentic AI is already yielding measurable business outcomes in the Silicon Savannah. For Savings and Credit Cooperative Organizations (SACCOs) and microfinance institutions, the technology solves the most pressing pain point: the high administrative cost of processing low-margin micro-loans.
By deploying autonomous agents to handle the verification and underwriting processes, institutions can scale their lending operations exponentially without proportionally increasing their human headcount. Furthermore, these agents operate without fatigue or bias, ensuring that regulatory compliance checks are executed flawlessly on every single transaction, insulating the institution from punitive regulatory fines.
Global Parallels in Algorithmic Governance
The push by firms like Akili AI to institutionalize AI governance in East Africa closely mirrors the regulatory shockwaves currently hitting London and Brussels. The European Union’s sweeping AI Act imposes draconian penalties for the negligent deployment of autonomous systems in high-risk sectors like banking and healthcare.
For Kenyan enterprises harboring global ambitions, or those seeking foreign direct investment from UK or US private equity, demonstrating rigorous AI readiness is no longer optional. Foreign investors now demand explicit “AI Risk Registers” as a standard component of due diligence. Akili AI’s emphasis on MVP-first deployment and strict guardrails provides a vital compliance blueprint for African tech firms navigating the complex web of international data sovereignty laws.
The mandate is clear: African businesses must stop playing with generative chatbots and begin the serious, structural work of integrating autonomous agents into the bedrock of their operations.