AI capabilities are advancing faster than many organizations are ready to operationalize and monetize them. While soaring AI investment reflects high value creation hopes, it also raises expectations and shortens patience for strategies that don’t translate into results. That’s a quiet tension facing many executives.
The gap shows up clearly in KPMG’s Q1 2026 AI Quarterly Pulse Survey of 237 US senior leaders. Average projected AI investment has nearly doubled year-over-year, from $114 million to $207 million. Yet, 65% cite difficulty scaling use cases (up from 33% last quarter) and 62% point to skills gaps (up from 25%) as the top ROI. Exacerbating the challenge, traditional ROI models often lag productivity and performance gains, clouding value when leaders most need confidence.
Those organizations already realizing AI’s value effectively and securely reengineer processes and operating models at enterprise scale. For others still chasing results, upping spending — without substantively addressing AI strategy, governance, workforce capabilities and cross-functional collaboration — burns cash, impairs competitiveness and erodes morale.
Credible c-suite responses to four questions can change that.
1. Does our AI investment drive new revenue — or merely defend margins?
Customer experience (64%) ranked as survey respondents’ top investment priority, slightly topping analytics and cybersecurity. That’s quite telling about AI investments’ strategic intent for most — building and monetizing new products and services. Sales expansion lacks cost reduction’s inherent limits and that distinction compounds over time.
KPMG principal and aIQ program lead Rahsaan Shears identified a scaling trend in the data, explaining, “The biggest shift isn’t belief in AI — that debate is largely settled — it’s confidence grounded in results. Leaders are now seeing tangible business value in productivity, speed, and decision quality, and that proof is accelerating investment. What’s changed is the move from experimentation to scaling what’s already working.”
The data reflects that conviction. Seventy-nine percent of leaders say AI will remain a top investment priority even facing economic downturn. That signals betting on growth, not the typical posture of enterprises protecting margins.
2. Is governance built into our AI systems from the start — or only added after something goes wrong?
Almost all (91%) of the surveyed executives cited data security, privacy and risk concerns as the top factor shaping their AI strategy. Forty-four percent are deploying AI agents from trusted technology providers or have ring-fenced high-risk use cases where autonomous decision-making is not permitted. Another 43% design monitoring and evaluation controls directly into agent architecture. There’s a sizable execution gap between “built-in” rather than “bolt-on” governance.
The human-in-the-loop trajectory tells the story. Adoption has climbed from 32% to 44% to 57% across three consecutive quarters, but Shears urges a closer look. “The rising adoption rate is encouraging, but it reflects very different levels of maturity. Forward-looking organizations are designing governance, oversight, and escalation into AI systems from the outset. Others are still responding defensively — adding human review only after issues surface, rather than embedding it as a core part of how AI operates,” she said.
Governance after deployment is far costlier than by design. “External cyber threats and internal misuse converge when access, visibility, and accountability break down. The leaders who will succeed are investing early in Trusted AI — putting secure, approved tools in the hands of their people, backed by a strong governance layer, centralized orchestration, and a single control plane — so innovation happens inside the guardrails, not around them, as agents become part of daily work,” Shears emphasized.
3. Is AI embedded across functions — or does it reinforce longstanding silos?
Many organizations have deployed AI broadly. Very few have connected it.
More than half are now actively scaling or deploying AI agents — a tipping point the report flags as significant. Among those deploying, 73% report AI agents automating workflows that span multiple functions, 53% are routing information and decisions between teams, and 51% have deployed shared knowledge bases or unified dashboards. Those numbers suggest progress — but 65% simultaneously report difficulty scaling use cases and 36% struggle to move AI across teams and functions.
Deployment and orchestration are starkly different. Business models reliant on entrenched siloed workflows cannot easily adapt to AI systems built to span functions and spur coordination. Pushing AI without careful workplace dynamics consideration may have the unintended downside of hardening workplace fiefdoms. That’s how scaling backfires.
4. Is AI capability spreading across the enterprise — or concentrated in specialist teams labeled digital strategists?
Centralizing AI capability in a specialist team creates an illusion of progress. It concentrates expertise rather than distributing it, leaving the broader organization unable to execute at scale. With skills gaps spiking from 25% to 62% in a single quarter, that illusion is becoming harder to sustain.
Organizations are responding by upskilling and reskilling (87%), recruiting for AI-native roles (68%) and redesigning jobs entirely (55%). Willingness to pay a premium for AI skills has jumped sharply as well, with 45% (up from 22% in Q4) now willing to pay 11-15% more.
“Organizations making progress recognize that technical skills alone aren’t enough. They’re pairing technical AI training with capabilities like critical thinking, judgment, and problem solving — skills that help people reimagine how work gets done with AI. Instead of generic training programs, they’re equipping employees to think differently, ask better questions and apply AI meaningfully in real work,” Shears observed
Recruiting expectations have shifted to match, as AI agents have already changed respondents’ approach to entry-level (64%) and experienced (71%) hiring. When evaluating AI-era talent, 83% of leaders now prioritize adaptability and continuous learning — outranking technical or programming abilities, cited by 71%.
“Employers should hire for learning capacity, adaptability, and judgment. For candidates, the differentiator isn’t how fast you can work without AI, but how effectively you can think, collaborate, challenge it to enable better decision making alongside it,” Shears advised.
Winning Ways
Organizations unlocking the most value from AI pair ambition with accountability. They invest not only in technology, but in strategy, governance, workflow redesign and talent at scale. Otherwise, bigger budgets simply make the gaps pricier.
The key question is well past whether to invest in AI, but who’s driving value?
