Employees are increasingly using unapproved AI applications for routine tasks such as summarising meetings, drafting documents and analysing data within Kenyan businesses.

A poll conducted during a keynote by Akili AI chief executive Dr Simon Bransfield‑Garth at the Safaricom Cybersecurity Summit 2026 revealed that among roughly 150 attendees, 42% confirmed they knew unapproved AI was being used in their organisations, while another 25% believed it was almost certainly happening.

Only 6% of respondents said their company had a fully developed AI strategy and governance policy that had been reviewed and supported by staff training.

The poll reflects the views of summit participants rather than a statistically representative survey of Kenyan businesses.

Unapproved AI tools can expose confidential information when employees submit names, financial details or internal plans to external services without a formal data‑sharing arrangement.

The risk varies with the specific tool, its configuration, the type of data sent and the organisation’s existing contractual and technical safeguards.

Blanket bans on AI services are hard to enforce and may drive staff to seek unsanctioned alternatives, while also preventing the use of tools that could improve efficiency and customer service.

Dr Bransfield‑Garth recommended a four‑part approach—measure, manage, train and maintain evidence—to bring shadow AI under control.

Measurement involves identifying all AI tools employees use, including embedded features in existing software.

Management requires clear acceptable‑use policies, risk assessments and incident‑response procedures, while training helps staff recognise data‑exposure risks.

Companies should assign responsibility for AI governance across legal, compliance, security, HR and operations rather than leaving it solely to IT or individual employees.

The immediate step for businesses is to inventory current AI usage and assess what data is being processed, enabling proportionate controls without treating every AI application as either harmless or prohibited.