Cybersecurity Experts at CAISEC Call for Stronger AI Governance and Data Protection Amid Rapid Rise of AI Agents
Cybersecurity and artificial intelligence experts participating in a panel discussion during CAISEC 2026 stressed the urgent need for stronger governance frameworks, data protection measures, and cybersecurity controls to keep pace with the accelerating adoption of AI agents across organizations and industries.
The session, titled “Securing Agentic AI: When AI Works on Your Behalf,” highlighted the growing challenges associated with deploying autonomous AI systems capable of making decisions, accessing sensitive information, and executing operational tasks with limited human intervention.
Speakers emphasized that organizations must adopt balanced strategies that maximize the benefits of AI technologies while ensuring the security of institutional systems, digital identities, and sensitive data.
Sherif Sheltout, Regional Executive Director for the Middle East and North Africa at Liquid C2, said global indicators show a sharp acceleration in enterprise adoption of AI agents, creating new cybersecurity, governance, and risk management challenges.
According to Sheltout, data from BeyondTrust revealed that AI agent adoption rates increased by 466 percent year-on-year, while Veeam estimates that some enterprise environments now operate with as many as 82 AI agents per employee.
He also pointed to studies showing that 97 percent of organizations maintain excessive or unnecessary access privileges within their digital environments, increasing the risk of misuse, insider threats, and cyberattacks.
Sheltout added that Cloudflare data indicates that 57 percent of global web traffic is now generated by bots and automated systems, while Cisco reported that 85 percent of organizations are currently testing Agentic AI applications, despite only five percent deploying such systems into full production environments.
“These figures highlight a clear gap between the speed of AI adoption and the development of corresponding governance and security frameworks,” Sheltout said, calling for stricter controls to ensure the safe and responsible use of AI technologies.
Stefano Maccaglia, Director of Incident Response and Red Team Operations at NetWitness, stressed the importance of clearly identifying the nature and operational role of AI applications before granting them extensive access privileges inside organizations.
He warned that malicious actors could increasingly exploit AI agents to automate harmful activities and accelerate cyberattacks, noting that tasks previously requiring significant time and technical effort can now be executed much faster using AI-powered tools.
Maccaglia emphasized the need for continuous testing and behavioral monitoring of AI agents to ensure secure data handling and prevent misuse, particularly given the limited historical experience and reference frameworks surrounding these emerging technologies.
Mohamed El-Shall, Strategic Accounts Manager at Cloudflare, highlighted the importance of securing data flows and preventing unauthorized usage of enterprise information.
He stressed the need to distinguish between Generative AI systems and Agentic AI technologies, especially regarding data access privileges and operational autonomy.
Hassan El-Bamby, Country Manager for Egypt, Bahrain, and Jordan at BeyondTrust, said the unprecedented growth in AI adoption requires organizations to implement stronger governance and oversight mechanisms.
Meanwhile, Mostafa El-Anany, Country Manager for Egypt and Libya at Veeam, emphasized that organizations must begin by classifying data according to sensitivity and defining clear ownership and access responsibilities before allowing AI systems to interact with critical information.
Ibrahim Shamel, Head of Cybersecurity Sales at Cisco Egypt, stated that organizations should first determine the business value AI technologies can deliver before focusing solely on deployment or security mechanisms.
“The success of AI adoption depends on clearly defined objectives and operational roles, not merely following technology trends,” he said.
Hazem Moftah, Sales Engineering Manager at Forcepoint, described data governance as the foundation for secure AI adoption, warning that organizations face growing challenges related to accountability, access management, and sensitive information protection.
He added that cybersecurity in the AI era has evolved into a broader operational culture focused on understanding data context, controlling access, and making rapid decisions capable of keeping pace with fast-moving technological developments.
Participants agreed that successful AI adoption requires integrated governance models covering identity management, access controls, data protection, and operational accountability to ensure organizations can safely benefit from advanced AI capabilities while minimizing associated risks.
The discussions formed part of CAISEC 2026, one of the region’s leading cybersecurity and digital sovereignty events, which this year focused heavily on AI-powered threats, cloud security, digital governance, and the future of intelligent autonomous systems.