
Kenyan businesses are rapidly embracing artificial intelligence (AI) to improve productivity and automate operations, but lack of safety structures is still a major issue.
However, weak governance structures, inadequate employee training and limited investment in risk management are exposing organisations to potential data breaches, regulatory penalties and reputational damage.
A new report by Akili AI and the Kenya National Chamber of Commerce and Industry (KNCCI) shows that 99 per cent of organisations surveyed are already using AI in some form, ranging from publicly available tools to artificial intelligence features embedded in everyday business software.
However, only about one in four organisations has integrated AI into standard workflows across the organisation, highlighting the gap between widespread adoption and structured implementation.
The report, titled AI Readiness in Kenya 2026, warns that employees are increasingly introducing AI tools into workplaces without formal approval, leaving businesses with limited visibility over the technology being used and the information being shared with external platforms.
Akili AI founder and chief executive Simon Bransfield-Garth said AI adoption in Kenyan workplaces had largely happened without deliberate organisational planning, exposing companies to risks involving privacy, data management, decision-making and cybersecurity.
The findings are based on 214 responses collected in October 2026 during the KNCCI Africa Commerce and Industry Summit and the Safaricom Cybersecurity Summit.
“Four in ten organisations surveyed are already using AI in core business processes or customer-facing services, including finance, human resources, credit assessments, insurance claims and customer service,” said Bransfield-Garth.
Yet three-quarters of the organisations have no AI policy in place, while four in five lack a designated executive responsible for managing AI-related risks.
The governance gap is particularly concerning as employees increasingly rely on publicly available AI applications to generate reports, process information and perform routine tasks.
The report found that between 61 and 69 per cent of respondents either knew employees were using unapproved AI tools or had no way of establishing whether such applications were being used.
This growing use of unapproved applications, commonly referred to as shadow AI, raises concerns about employees sharing confidential company information and personal customer data with platforms whose data storage and processing arrangements they may not understand.
“Businesses cannot effectively manage risks associated with technology they cannot identify, companies should establish inventories of AI applications being used across their operations,” the report states.
Data protection compliance is another major concern, particularly as businesses increasingly deploy AI in decisions affecting customers and employees.
Only about one in six organisations has conducted a documented AI impact assessment, despite existing obligations under the Data Protection Act, 2019.
The law already applies to personal information entered into AI applications, including data transferred to offshore service providers and decisions made through automated processing.
The Office of the Data Protection Commissioner has also published AI-specific guidance outlining requirements for organisations developing, deploying or procuring AI systems that process personal data.
The report notes that approximately seven in ten organisations would either fail to detect an AI-related incident or lack a defined response mechanism.
This leaves businesses exposed to potential data protection violations, cybersecurity incidents and reputational damage as AI becomes increasingly embedded in their operations.
Beyond governance, inadequate investment in skills and infrastructure is threatening the ability of businesses to extract meaningful value from the technology.
The survey found that 63 per cent of organisations provide no structured AI training, leaving employees to learn through personal experimentation or informal guidance.
Leadership understanding is also limited, with 58 per cent of business leaders surveyed at the KNCCI summit indicating that senior management either does not understand AI or relies entirely on others to make decisions about the technology.
Investment remains another hurdle, with 68 per cent of organisations surveyed at the cybersecurity summit and 75 percent at the KNCCI summit lacking a defined AI budget.
Only 10 per cent and 19 percent respectively had budgets covering AI security, governance and employee training, rather than technology tools alone.
“A bout 94 per cent of cybersecurity summit respondents and 98 per cent of KNCCI summit respondents view AI as more of an opportunity than a risk,” the report states.
The report recommends that businesses begin by identifying AI tools in use, appointing accountable executives, establishing clear usage policies and training employees on responsible adoption.