The United Arab Emirates continues to face a mounting wave of cyberattacks, with reports indicating the country experiences over 800,000 attempted breaches daily. This surge, highlighted during Cybersecurity Awareness Month in October, underscores growing concerns about the evolving threat landscape, particularly the role of autonomous artificial intelligence (AI) in facilitating cybercrime.
According to the UAE Cyber Security Council, daily intrusion attempts targeting the country’s infrastructure escalated from between 90,000 and 200,000 earlier this year to surpass 800,000 in recent months. The council has documented 128 confirmed incidents involving ransomware, data leaks, and website defacements. Regional cyberattack analysis from Positive Technologies shows that the UAE accounts for approximately 35 percent of cyber threats across the Gulf Cooperation Council (GCC) region.
Experts attribute much of this surge to the increasing use of autonomous AI agents, which operate with limited human supervision and are capable of independently making decisions, accessing data, and navigating corporate systems. Dr. Mohamed Al Kuwaiti, head of the UAE Cyber Security Council, noted that AI has significantly accelerated both the scale and sophistication of cyberattacks globally. Cybersecurity losses reached an estimated $10.5 trillion in 2024 and are projected to rise to $15.6 trillion by 2029.
Industry leaders emphasize that these AI-driven attacks differ fundamentally from previous threats. Salah Suleiman, Managing Director for South Gulf at TrendAI, explained that each autonomous AI agent effectively creates a new identity and attack vector. This shift enables criminals to execute faster, more convincing, and scalable operations. In response, TrendAI advocates for a security framework based on visibility, intent, control, and oversight to manage the behavior of AI models and agents from the outset.
Governance remains a critical challenge. Alexander Depret-Bixio, Senior Vice President at Anomali, emphasized the importance of defining parameters around AI decision-making, particularly given the risks tied to poor data quality. “Confident but bad decisions” by AI agents represent an unmanageable risk in cybersecurity environments.
Research by DXC Technology reveals that less than half of organizations globally—47 percent—have fully adopted governance structures for agentic AI, even as many deploy these autonomous systems. Daniel Chuckes, DXC’s cybersecurity director for Asia Pacific and Japan, warned that companies expanding AI capabilities without corresponding governance measures are effectively scaling risk in tandem with their technology.
Experts advocate a shift from reactive to proactive cybersecurity strategies that integrate intelligent, autonomous defenses governed by stringent human oversight. Justin Henkel, Chief Information Security Officer at SolarWinds, frames autonomous machine identities as an enterprise control issue, requiring clear purpose, authority limits, provenance, behavioral transparency, and revocation processes. He emphasized that managing these identities demands measurable, ongoing discipline rather than one-time compliance efforts.
As cybercriminals increasingly leverage AI to exploit organizational weaknesses, security specialists call on corporate boards to actively determine the extent of autonomy granted to AI systems and to establish mechanisms for auditing every decision. According to Depret-Bixio, the credibility of AI’s near-term future hinges on the ability to “stand behind every decision it makes,” marking a critical juncture in the evolution of cybersecurity governance worldwide.