Check Point Software Technologies recently launched a “Secure AI Advisory Service” to help companies manage and accelerate AI adoption while ensuring that governance and oversight keep up.
The move comes as research shows AI adoption in South African companies is accelerating faster than the policies and controls needed to manage it.
The South African Generative AI Roadmap 2025 by World Wide Worx for Dell Technologies and Intel reports 67% of South African companies now use GenAI, up from 45% in 2024.
However, the study also states that formal governance lags adoption. Only 14% of companies have a company-wide AI integration strategy, while 32% report using AI tools unofficially.
Shayimamba Conco, cyber security expert at Check Point, said a number of companies are deploying AI systems faster than they can secure or govern them.
“Many organisations are adopting AI rapidly to remain competitive, but governance and security frameworks are still catching up,” he said. “It’s important that AI deployments follow a security-by-design approach, where risk management and data protection are built in from the start.”
According to Conco, companies often underestimate the risks associated with AI systems when these are introduced.
“Common gaps include lack of visibility into how AI models access sensitive data, weak access controls and limited policies governing AI usage. Organisations also often overlook risks like data leakage, prompt injection and model manipulation,” he said.
According to Check Point, the new service is designed to help companies address these gaps by setting governance frameworks, assessing risks and aligning AI systems with international regulations.
International rules are also expected to shape how companies manage AI, even outside the regions where they originate.
“Global regulations like the EU AI Act will influence African organisations, especially those doing business internationally. Even where compliance isn’t mandatory, these frameworks often set the benchmark for responsible AI governance and security practices,” Conco added.
The World Wide Worx research suggests many companies still have significant work to do. It found 37% of companies using AI have no safeguards in place for responsible use, while only 13% report having comprehensive guardrails covering privacy, safety and bias risks.
Despite these challenges, businesses remain optimistic about the technology’s potential. The study found 70% of companies expect AI to deliver significant or major improvements in growth, innovation and operational efficiency.
Check Point said its Secure AI Advisory Service aims to help companies close the gap between adoption and governance by providing risk assessments, policy frameworks and guidance on implementing security controls.
Conco said companies should start with clear governance policies, strong identity and access controls, and ongoing monitoring of how AI systems interact with sensitive data.
“A prevention-first security approach is key to protecting both AI models and the data they rely on,” he said. “AI brings enormous opportunity, but security and governance must evolve alongside innovation.”
Share