Kathy Gibson reports from AI Everything x Gitex Kenya – Africa is once again showing that it’s not afraid to innovate in a world of rapidly changing technology.
Gitex Kenya last week showcased the energy that the region is known for, with a thriving ecosystem of enterprises, developers and startups all keen to seize new opportunities.
“East Africa is booming for us,” says Chris Norton, GM of Kaspersky Africa. “It’s not alone: West Africa and North Africa are both doing very healthy business too.
“But we are doubling down on the East Africa region, and we are planning to make some significant investments.”
Norton points out that, while artificial intelligence (AI) is all the rage at the moment, organisations shouldn’t lose sight of the myriad other threats they could face. And, in some instances, East African users have moved so quickly embrace new technologies that they might have missed some vulnerabilities.
“This region is obviously keen on AI at the moment, but it’s not the only thing organisations should be thinking about. AI must be used appropriately, and employed to solve the problems that it is simultaneously creating in the cybersecurity environment.”
AI-based attacks are more frequent and more virulent, Norton says. “But while AI is getting the attention, many enterprises in this region haven’t yet started protecting against things like holomorphic or polymorphic attacks.
“So there is a whole learning curve still ahead of the industry in this region.”
Kaspersky data demonstrates that in 2025, password stealer attacks increased by 83% year-over-year in Kenya and 56% across Sub-Saharan Africa. Spyware attacks grew by the same figure of 83% in Kenya and 53% regionally, while backdoor attacks rose by 25% in Kenya and 8% across Sub-Saharan Africa. Although exploit attacks showed a slight decline, they remain a major concern due to their mass spread and unauthorised access they open to a users’ systems. Meanwhile, ransomware continues to pose a serious risk to organisations, with 7.62% of organisations in Africa experiencing ransomware detections in 2025.
Advanced Persistent Threats (APTs) remain among the most serious risks for enterprises. According to the Kaspersky Security Services Global Report, APT groups were detected and blocked in 21% of customers in 2025 and accounted for 23% of all high-severity incidents. These highly organised groups increasingly combine AI-enhanced techniques with social engineering and targeted intrusion methods to maximise operational effectiveness.
On the positive side, he adds, companies in Africa will be able to leapfrog generations of technology. So they’ll be able to go from the initial phases of AI straight into advanced AI concepts and capabilities without having to go through the same learning curve.”
At the same time, he worries that governance might be lagging. “Who is putting the guardrails in place and who is monitoring them?”
Just as Shadow IT plagued CIOs in previous decades, he points out that Shadow AI has the potential to wreak havoc among organisations’ cybersecurity.
“This is where employees bring AI tools into the corporate environment and connect them to the internal business systems. They are giving these AI tools access to company confidential information that’s then going out into the wild, and will become generally available in those tools.”
In South Africa most enterprises, and all of the big banks, limit access to their corporate intelligence, but some of the Africa-based organisations haven’t closed this loophole yet, Norton says.
All in all, Norton sees palpable excitement and energy in the market, and believes the uptake of AI will be fast. Kaspersky has had many years of experience in the AI space.
According to Kaspersky experts, cybercriminals can use AI across multiple stages of cyberattacks: from preparation and communication to assembling malicious components, probing for vulnerabilities and deploying tools, while simultaneously concealing evidence of AI involvement to complicate investigations and attribution. Malicious actors are also actively distributing malware disguised as AI tools to steal sensitive information from victims.
One of the growing cybersecurity issues is the spread of deepfakes and AI-generated fraudulent content. As AI tools become more and more sophisticated, distinguishing authentic material from manipulated ones is becoming more difficult. Kaspersky researchers warn that AI models can also be vulnerable to “unintended memorisation”, where models retain fragments of sensitive information that attackers may later extract. Additional risks include malicious tampering with training datasets, injection of harmful logic into AI software code and exploitation of vulnerabilities within AI-powered systems.
The emergence of AI agents, which are systems capable of autonomously taking actions on behalf of users, creates another significant attack surface. According to Kaspersky, these systems can be manipulated through adversarial content or misconfigured autonomy settings, potentially leading to harmful real-world actions.
Kaspersky recommends organisations to regularly assess AI-related risks and establish comprehensive AI governance policies defining which AI tools are approved and what types of data can be processed. Regular employee training on secure AI usage, recognition of fake AI services, malicious links and prompt injection risks is equally essential.
For private users, Kaspersky recommends exercising caution when using AI-powered tools, carefully reviewing privacy settings, verifying the authenticity of AI applications and double-checking information generated by agentic AI systems before making decisions based on automated outputs. The company also advises families to maintain open discussions with children regarding their use of AI technologies and online safety practices.