Microsoft Corp. is expanding a specialized, multi-model artificial intelligence program to South African enterprise clients, aiming to counter a surge in highly sophisticated, AI-driven cyberattacks targeting the continent’s largest economy.
The system, codenamed MDASH, utilizes an orchestra of more than 100 individual AI agents that draw from a mix of major frontier models and smaller, distilled algorithms. Developed by Microsoft’s Autonomous Code Security team, the program automatedly discovers, validates, and patches code vulnerabilities end-to-end. The rollout marks a critical shift for the tech giant, transitioning a tool initially built to safeguard its own proprietary source code into a private usage defense mechanism for external corporate networks.
The deployment comes amid a stark warnings over a growing defensive deficit in Africa. A recent Boston Consulting Group report found that while nearly 60% of African companies suffered AI-enabled cyberattacks over the past year, only 29% have deployed advanced AI security tools to defend themselves. Furthermore, just 3% of organizations on the continent have significantly increased their cybersecurity budgets in response to the escalating threat.
“Organisations can no longer rely on legacy approaches to security alone,” said Kerissa Varma, Microsoft’s chief security advisor for Africa, in an interview. “In a world where vulnerabilities can be discovered and exploited at scale, manual processes simply cannot keep up. The need for modern, cloud-based infrastructure and AI-driven defence has become urgent.”
The technical barrier to entry for cybercriminals has plummeted as commercial AI models match human experts in spotting digital flaws. According to Microsoft’s latest Digital Defence Report, AI-generated phishing campaigns are proving 4.5 times more effective at tricking corporate employees than traditional scams. The threat vector is particularly acute in South Africa, where a expanding digital footprint intersects with systemic tech budget constraints and an acute shortage of specialized tech talent.
Unlike conventional corporate security programs tethered to a single software provider, MDASH selects different AI models based on the specific computing task, matching raw processing power with cost efficiency. The platform recently ranked first on the CyberGym AI security benchmark, outperforming single-model setups.
“Cost will be an important consideration as businesses look to scale AI securely and sustainably,” Varma said. “The good news is that some of the most cost-effective models are exceptionally effective at specific security tasks, it’s about using the right tool for the right job.”
Microsoft is increasingly leaning on AI-assisted vulnerability findings to feed its global “Patch Tuesday” software updates, warning that the volume of discovered threats is rising month-on-month. The company is urging local firms to migrate away from legacy, on-premises servers toward automated cloud architecture to keep pace with the patching cycle.
To mitigate the regional talent crunch where 82% of African organizations report severe difficulties hiring AI-cybersecurity professionals, Microsoft is expanding a skills campaign across South Africa’s 50 technical vocational colleges via a partnership with Ikamva Digital.
The tech giant is also widening access to its “Secure Now” framework, offering security advice to non-Microsoft customers as part of its broader Secure Future Initiative.
“The conversation should not be about access to a single model,” Varma said. “It’s about building a comprehensive, resilient security strategy with the right capabilities in place today.”
