Professional-services firms are finding that faster AI-assisted tasks do not automatically produce faster client work. A SiliconANGLE interview points to a telling gap: 12% of chief executives surveyed said AI had brought both revenue growth and cost reductions, while 56% reported no significant financial benefits.
Watch Desk analysis
What happened
PwC U.K.’s Matt Cook and Certinia’s Prasad Narasimhan Sulur told theCUBE that firms need to redesign whole workflows around AI, rather than simply bolt tools onto existing processes. Sulur gave software development as an example: AI may shorten coding, but testing, integration and release can still hold up delivery. In consulting, he said, AI could help compress research and analysis from months into weeks.
The article attributes the survey figures to PwC’s 29th Global CEO Survey. It also notes that the interview segment was sponsored by Certinia.
Why it matters
The figures suggest a useful distinction between making individual tasks quicker and changing how an organisation delivers work. If the rest of a process remains slow, automating one step may produce a better demonstration rather than a better business result.
The interview also describes the less glamorous work that makes AI useful in practice: giving systems relevant business context, managing permissions and checking their outputs. Cook’s point is that AI can accelerate evidence gathering, but professional accountability remains with people.
Our read
This is a more grounded AI productivity argument than another race to count generated lines of code. The measure that matters is whether a redesigned process improves the outcome, not whether one task looks impressively quick. The survey numbers are a useful warning against treating adoption as proof of payoff.
What to watch
- Whether companies report measurable gains from redesigned workflows, not just faster individual tasks.
- How firms provide AI systems with useful context while controlling access to sensitive information.
- Whether human review remains proportionate as AI takes on more routine work.
Discussion spark: When AI speeds up one step but leaves the rest of a process unchanged, should that count as a productivity win?
Sources and evidence
- Human judgment reshapes AI services – SiliconANGLE (10 October 2026, 20:22 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.