Short-term risk from AI and cyber security has spiked, warns Marsh Africa

Marsh Africa has warned that short-term risks associated with AI and cyber security have surged, as South Africans are adopting AI at a pace faster than institutions can respond. The company's research, based on the Oliver Wyman Forum's 300,000 Voices study and a global consumer survey, reveals that AI adoption is outpacing the capacity of organizations to manage the associated risks.

Ben Wilmot-Sitwell, Marsh South Africa's cyber growth leader, stated that the growing cyber threat landscape is directly linked to insurance investments and claims. The study also highlights phishing as a persistent challenge, with generative and agentic AI making phishing campaigns more sophisticated. Cybersecurity expert Spiros Fatouros stressed that business leaders are increasingly concerned about the risk and its potential impact on their clients' ability to detect, respond to, manage, and contain cyber incidents.

Risz Moola, founder and CEO of HyperionDev, warned that AI-enabled attacks are threatening finance, HR, customer service, and other non-technical teams. He attributed this to AI's rapid adoption, making it increasingly difficult for organizations to recognize and mitigate such threats. Additionally, AI usage at work has increased significantly, with daily use rising from 22% in 2023 to 26% in 2025, and 60% of South Africans reporting improved work-life balance due to AI.

However, concerns about job security persist, with 61% of employees believing that over half of their company's roles could be replaced by AI.