Organizations are accelerating artificial intelligence (AI) deployment despite known security and compliance risks, according to a new global study by TrendAI.

The survey of 3,700 IT and business decision-makers, including 200 from the Philippines, found that 67 percent of IT leaders and 72 percent of business leaders locally felt pressured to approve AI projects despite concerns.

One in seven described those risks as “extreme,” but still moved forward to keep pace with competition and internal demand.

While most organizations have measures in place to protect data, including security controls and quality management systems, the study found that AI adoption is outpacing governance.

Only a third of IT leaders and less than half of business leaders said they feel highly prepared for the speed of AI deployment.

TrendAI said gaps in oversight are being worsened by unclear accountability and inconsistent governance frameworks, with some firms resorting to unsanctioned or “shadow AI” tools.

The risks are growing as cybercriminals begin to use AI to automate attacks, including phishing and reconnaissance.

Organizations also flagged concerns over autonomous or “agentic” AI systems, particularly their access to sensitive data and lack of visibility once deployed.

In the Philippines, 65 percent of respondents support introducing an AI “kill switch” to shut down systems in case of misuse or failure.

“Organizations are not lacking awareness of risk, they’re lacking the conditions to manage it,” said Rachel Jin, chief platform and business officer at TrendAI.