Safaricom Cybersecurity Summit Flags Rising AI Threats

In Short

Safaricom held its fourth Cybersecurity Summit in Kenya under the theme “Harnessing AI, Strengthening Security.” The focus was on AI-driven cyber risks and practical steps to improve cyber resilience.

What Happened

Safaricom brought together players across Kenya’s digital ecosystem for a one-day cybersecurity summit aimed at improving cyber resilience, which means an organisation can prevent attacks, keep operating during incidents, and recover quickly.

A key theme was the “dual use” nature of AI, meaning the same tools can help defenders spot attacks faster, and also help criminals scale fraud. Safaricom highlighted threats such as deepfakes, which are AI-made fake audio or video that can mimic real people, AI-generated impersonation, automated scams, and ransomware, which is malware that locks your data until you pay.

Safaricom cited data from the Communications Authority of Kenya showing the National Kenya Computer Incident Response Team detected 11.12 billion cyber threat events in the last financial year, up 29% from 8.62 billion the year before. A “threat event” can include scans, probes, and attempted attacks, not only successful breaches.

During sessions, participants discussed AI-enabled threat detection, incident response planning, ransomware readiness, and the cybersecurity skills gap. Safaricom also pointed to security culture, meaning staff are trained to spot unusual activity and verify high-risk requests, as a key part of resilience.

Why It Matters

Kenya’s economy is increasingly digital, and more services now rely on apps, APIs, and mobile money rails. As AI makes fraud easier to automate, businesses may need to move security earlier in the product lifecycle, not treat it as a final check.

Safaricom also positioned micro, small, and medium enterprises as a priority, since they power a large share of GDP but often lack in-house security teams. The telco said it supports MSMEs through SD-WAN, which is centrally managed secure connectivity across branches, and a Managed Security Operations Centre, which provides continuous monitoring and coordinated response.

For founders and operators, the message is clear, plan for identity fraud, social engineering, and ransomware as standard risks, and test your response playbooks before an incident happens.