From Pilot to Pause: Why GenAI is Stalling

Gartner expects that 30 percent of GenAI projects will be abandoned by the end of this year. Why? Because the fundamentals don’t hold up. Poor data quality. Loose risk frameworks. Sky-high costs and unclear outcomes.

Some companies are spending up to $20 million just to test something they can’t justify keeping.

At the same time spending is ballooning. Gartner and IDC forecast global AI investment will hit $644 billion this year. That’s up more than 75 percent from last year. But here’s the catch. Most of it is on hardware. That tells us this is still a capability race. Not a value race.

In private most CEOs are now asking a different question. Not “How can we use AI?” but “What are we getting from this?” And the answer is not always pretty.

The Wall Street Journal reports that about 70 percent of GenAI projects are stuck in pilot mode. The tools just aren’t accurate enough yet to trust in production.

The bubble word is coming back. Sequoia and MIT are warning that the GenAI boom may be outpacing real returns. Big Tech spent over $200 billion on it last year alone. The gap between hype and results is starting to worry even the believers.

So what does this mean for you?

1 - Ask sharper questions. Tie every AI project to a business objective. Then put a 12-month clock on it. If it doesn’t deliver, drop it.

2 - Check your data foundations. Bad data means bad AI. The faster you scale the more dangerous it gets if you don’t have controls in place. Put your smartest people on AI governance. You can’t afford reputational risk from a rogue deployment.

3 - Give your teams permission to kill projects that are dragging. Fast failure is better than slow waste.