The technology industry is locked in a high-stakes sprint toward autonomous artificial intelligence (AI), prioritizing market dominance over safety protocols even as internal oversight systems buckle under pressure.
According to the latest EY US Technology Pulse Poll, 97% of technology executives view the pursuit of autonomous AI as a “high” or “essential” priority for their long-term survival. However, this aggressive expansion has birthed what experts call a “velocity paradox,” where the speed of innovation is rapidly outstripping the ability to manage its consequences.
The survey of 500 U.S. technology leaders reveals a startling lack of institutional control. More than half (52%) of department-level AI initiatives are currently operating without any formal approval or oversight. Shadow AI in particular is already extracting a toll on corporate security.
In the past 12 months, 45% of executives reported confirmed or suspected leaks of sensitive data triggered by employees using unauthorized third-party generative AI tools. What is more, 39% reported similar breaches involving proprietary intellectual property.
“Technology companies continue to move at remarkable speed,” said James Brundage, EY’s global and Americas technology sector leader. “Without strong governance in place, companies risk hitting a plateau where large-scale growth becomes increasingly difficult. Leaders are currently navigating a balance between urgency and accountability.”
The report highlights a definitive shift in corporate philosophy. Rather than perfecting a product behind closed doors, 85% of tech leaders admitted they prioritize speed-to-market and iterative innovation, choosing to manage regulatory and ethical risks in real-time within a live environment. Only 15% of organizations still prioritize exhaustive pre-launch vetting.
Despite these growing pains, the appetite for investment shows no signs of waning. Nearly all (95%) executives plan to increase AI spending in the coming year, while 79% are specifically ramping up AI-related cybersecurity budgets, and 67% are increasing investments in cloud computing to support these systems.
Even though 70% of firms currently use a centralized model to manage AI approvals, the efficacy of these kill switches remains divided. Only half of the surveyed leaders say their AI governance or ethics heads have the independent authority to halt a high-priority, revenue-generating project that fails safety benchmarks. For the other 42%, stopping a risky project requires the direct intervention of the CEO or the board.
Beyond internal risks, external pressures are also mounting. 62% of leaders expressed concern that escalating geopolitical tensions and sovereign AI mandates could soon hinder their ability to scale these technologies globally.
“Organizations that standardize approved tools and strengthen monitoring will be better positioned to scale safely,” said Ken Englund, EY’s Americas technology sector growth leader.
For now, however, the industry appears content to build the plane while it is already in the air.
