AI in 2025: The Executive Reality Check
AI is here, creeping into your business, like it or not. Boards are expecting a plan, vendors are charging in, and the expectation sets the scene for missteps that will burn through cash quickly ahead of visible results. So, how do I look at this?
At FromHereOn, we see through the fog and make it easy to level the discussion to the issues that matter. In November 2024, we gathered 19 organisations across industries to review real-world AI strategy, governance, and adoption. The result? A no-nonsense view on what AI leaders are dealing with.
Want the full report?
We’ve distilled the key insights, real-world case studies, and AI leadership strategies from the FromHereOn AI Roundtable 2025.
Here’s what we learned
-
Executives want a sensible approach to AI. The problem? Problem: They struggle to define what they truly need. Many believe Copilot is “AI”, and that’s enough for the bean counters. Spoiler: It’s not. The real challenge is moving from AI dabbling to genuine capability building, which requires a more systemic view from strategy to execution.
-
Some AI tools aren’t even compliant with basic regulations, but that hasn’t stopped teams from pushing ahead. The governance debate is split: Some leaders are putting brakes on AI adoption out of fear, while others are freewheeling until legal catches up. Either way, the gap between AI ambition and AI reality is widening fast.
-
Once organisations go beyond entry-level use cases, they realise how quickly costs escalate. The astute leaders are tracking AI investments like an asset, not an experiment, tracking AI spending with a clear line of sight to business value.
-
Executives love AI—but not all problems need AI to solve them. Some of the biggest wins we heard weren’t from AI moonshots but practical applications of nascent AI capabilities in controlled environments that proved near-term impact. The real challenge isn’t deploying AI—finding the problems where a valuable and controlled approach to AI makes sense.
-
What is the most significant problem with AI adoption? People choose not to use it. Many leaders expect good AI use cases to support uptake, but even sensible AI rollouts can fail without effective engagement. Teams may resist AI because they don’t see what’s in it for them or how it may improve or even impact their jobs in the future. Want results? Solve the human problem first.
