Large businesses are increasingly anxious about disruption, but far less confident that AI will deliver the resilience they are publicly promising, according to new research from supply chain intelligence firm Zero100.
The study, which surveyed Chief Operating Officers (COOs) from $1bn+ companies, finds that cybersecurity is the dominant business continuity concern, while expectations for AI-driven transformation remain cautious and misaligned with what their CEOs are pitching to shareholders.
More than a third of businesses rank a cyber incident as the biggest threat to continuity over the next year (35%), ahead of geopolitical instability (20%), trade policy shocks (16%), and labour disruption (8%).
Cyber incidents stand out not just as the top concern, but as the fastest-moving shock companies expect to face. Nearly two thirds (62%) say they can respond to a cyber incident within minutes or hours, compared with disruptions such as tariffs, which take the vast majority of businesses days or weeks to respond to (83%).
However, opinion is sharply divided among COOs on whether AI will reduce or increase cyber risk (50% say better, vs. 43% say worse), even as a majority believe it will help manage supply shortages (64%) and ease skill gaps (58%).
And the research highlights a growing credibility gap between what companies are telling investors and what they believe they can deliver. Fewer than one in five COOs believe a majority of their company’s AI commitments to shareholders can be delivered on time (17%).
“Despite all the recent fallout from Davos, cybersecurity breaches rank as a much bigger business concern than geopolitical conflict,” said Lauren Acoba, VP, Research & Advisory Services at Zero100. “A major cyber failure can do far more immediate damage than tariffs or trade disputes to company profitability – and even entire economies, as last year’s JLR incident in the UK demonstrated.
“Furthermore, while CEOs are talking up AI to shareholders as a productivity engine, inside the business the mood is far more guarded. Operations leaders believe in AI’s potential, but they don’t believe the timelines.”
Expectations around agentic AI are particularly restrained. Just 7% believe agentic systems will fundamentally redesign a majority of workflows within the next two years. The most common view, held by two in five COOs (43%), is that agentic AI will reshape just 11–25% of workflows, pointing to selective deployment rather than wholesale transformation.
Progress on agentic AI is being slowed by organisational readiness rather than technical limitations. When asked specifically about deploying agentic AI at scale, COOs rated technology infrastructure as the most prepared area (6.2 out of 10), ahead of leadership understanding (6.0), data foundations (5.8), process maturity (5.6), and workforce skills (5.5).
Despite the promise of AI-enabled transformation, most company mindsets remain conservative when it comes to day-to-day priorities, limiting how much resilience AI can realistically deliver in practice. Keeping costs down matters more than anything else, accounting for nearly a third of COO performance metrics on average (29%), roughly double the weight given to growing revenue (15%) or improving customer service (17%). Longer-term signals such as brand trust barely feature at all, typically accounting for just 8%.
“Companies have the tools, but not yet the organisational muscle to change how work gets done using agentic AI,” Acoba added. “Cost control still dominates how they’re measured; they’re asked to chase growth, but are judged on profit. The result is an operating model that rewards not breaking things, more than trying new ideas. Until that changes, AI’s impact on resilience will be more incremental than transformational.”
The convergence of tech and brands 