TrendAI™, a global leader in AI security, has released new research revealing that organisations worldwide are continuing to accelerate AI deployment despite known security and compliance risks.

In Malaysia, the study found that 75% of IT decision makers and 70% of business leaders feel pressured to approve AI initiatives even when security concerns exist. One in seven respondents described these concerns as “extreme,” yet said they were still overridden in order to keep pace with competitors and internal demand.

AI adoption is clearly outpacing governance and control. Only 29% of Malaysian business decision makers reported being very or extremely confident in their understanding of legal frameworks governing AI, compared to 63% of IT decision makers. Around 52% of IT leaders said they already have comprehensive AI policies in place, while 39% are still developing them. Nearly half (51%) cited limited security or data expertise as the main barrier to safe AI adoption, followed closely by a lack of clear regulations and compliance standards (49%).


According to Goh Chee Hoh, Managing Director for Malaysia at TrendAI™, organisations in Malaysia are deploying AI faster than they can manage the associated risks, creating a widening gap between ambition and oversight. He noted that this reflects a broader disconnect in governance maturity among business leaders. To close this gap, leadership must move beyond reactive security and focus on full visibility into how AI interacts with data, supported by a centralised framework that embeds accountability and risk oversight into system architecture. He added that initiatives such as the MY-AI Standards are a positive step in addressing risks like deepfakes and algorithmic bias. However, as Malaysia moves towards its goal of becoming an AI Nation by 2030, organisations must strengthen governance and defensive resilience to ensure AI adoption leads to a sustainable digital future.


Rachel Jin, Chief Platform & Business Officer and Head of TrendAI, said that organisations are not lacking awareness of risk, but rather the conditions needed to manage it effectively. She explained that when AI deployment is driven by competitive pressure instead of governance maturity, systems are often integrated into critical infrastructure without sufficient controls, increasing exposure to risk. She emphasised that TrendAI remains focused on helping organisations achieve strong business outcomes while maintaining effective risk management.

The research also highlights that pressure-driven AI rollout is being worsened by inconsistent governance and unclear accountability for AI-related risks. Security teams are often forced to react to top-down AI decisions, which can lead to workarounds and increased use of unsanctioned or “shadow” AI tools. TrendAI’s latest threat research further shows that attackers are already using AI to automate reconnaissance, scale phishing campaigns, and lower barriers to cybercrime, increasing both the speed and scale of attacks.

Trust in autonomous AI systems remains at an early stage among business leaders. Fewer than half (44%) of Malaysian decision makers believe agentic AI will significantly improve cyber defence in the short term, citing concerns around data access, misuse, and lack of oversight.

More than half of organisations (57%) identify AI agents accessing sensitive data as their biggest risk. Around 45% are concerned about malicious prompts compromising security, while 37% fear misuse of trusted AI systems and risks linked to autonomous code deployment. At the same time, 36% admit they lack sufficient observability or auditability over these systems, raising concerns about how they can be effectively monitored or controlled once deployed.

Nearly half (48%) of organisations support the introduction of an AI “kill switch” to shut down systems in the event of failure or misuse, while more than half remain undecided. This highlights a broader lack of consensus on how to maintain control over increasingly autonomous AI systems.

“Agentic AI is moving organisations into a new risk category,” added Rachel Jin. “The concerns are already clear, from sensitive data exposure to loss of oversight. Without visibility and control, organisations risk deploying systems they do not fully understand or govern, and that risk will continue to grow unless action is taken.”

Read the full global report: Securing the AI-Powered Enterprise – Governance Gaps, Visibility Challenges and Rising Risk.

*TrendAI commissioned SAPIO Research to survey 3,700 IT and business decision makers across 23 countries globally.

About TrendAI™
TrendAI™, the global AI security leader and enterprise business unit of Trend Micro, empowers organisations with full AI visibility and unified security to drive innovation while reducing risk. Trusted by leading enterprises and governments across 185 countries, TrendAI™ secures organisations from identities to infrastructure and data.

Global Fortune 500 companies rely on TrendAI™ to reduce risk and detect threats up to three months earlier, powered by advanced threat intelligence.

Through ecosystem partnerships with NVIDIA, Anthropic, AWS, Google, and Microsoft, TrendAI™ helps organisations adopt AI securely and at scale.

AI Fearlessly.