AI needs process integration

Most big companies now have an artificial intelligence plan, but KPMG says only a small minority can show clear returns from it. (bankinfosecurity.com) KPMG’s Global AI Pulse, released March 31, 2026, surveyed 2,110 senior leaders in 20 countries and found 95% of organizations have an AI strategy. Just 39% said they are scaling AI across the enterprise, and 8% said they have seen tangible return on investment. (kpmg.com) (bankinfosecurity.com) Those companies still plan to spend heavily: KPMG put expected average AI spending at $186 million over the next 12 months. BankInfoSecurity reported the firms getting results are tying AI to operating metrics, governance, and day-to-day workflows instead of leaving it in isolated pilots. (kpmg.com) (bankinfosecurity.com) In plain terms, process integration means AI is built into the steps a business already uses to approve loans, handle claims, route service requests, or write software. It stops being a demo tool and becomes part of how work moves from one team to the next. (bankinfosecurity.com) (mckinsey.com) KPMG said the 11% it classifies as “AI leaders” share a few traits: they connect AI work to business results, use consistent performance measures across functions, and track impact while systems are running, not only after a project ends. Samantha Gloede, KPMG International’s global head of risk services and global trusted AI leader, said those firms also build “agent ecosystems” in an orchestrated way. (bankinfosecurity.com) (kpmg.com) The same survey points to the bottlenecks. KPMG said 58% of respondents are prioritizing information technology infrastructure spending, 50% are increasing cybersecurity and data-protection investment, and nearly three in four are concerned about data security, privacy, and risk. (bankinfosecurity.com) (kpmg.com) Talent is another divider. KPMG said leaders confident in their talent pipeline were four times as likely to report meaningful AI business value, 77% versus 20%, and said sustained training and change management are what let companies scale AI responsibly. (kpmg.com) That lines up with how consulting firms are framing the next phase of adoption. KPMG’s U.S. pulse in January said the leaders were “professionalizing” agents by preparing data, infrastructure, governance, and monitoring so systems can run reliably at scale. (kpmg.com) McKinsey made a similar case in banking in December 2024, arguing that material value comes when companies move beyond experiments and redesign critical workflows. Its examples included banks using AI in customer advice, loan-risk detection, and software development productivity. (mckinsey.com) The split KPMG describes is not between companies that bought AI tools and companies that did not. It is between companies still testing AI on the side and companies rebuilding the actual process around it. (bankinfosecurity.com)