“Using AI Isn’t Adopting It”: Our CEO on Kapital Empresa, on Moving from Pilots to Results

Madrid. One line from the interview sums up everything else: “using it isn’t adopting it”. Artificial intelligence is entering companies faster than organisations can digest it, and that gap was the subject of the conversation Francisco Moreno Balboa, our CEO and founder, had on Kapital Empresa, the daily business and entrepreneurship show.

The real impact hasn’t arrived yet — and the reason is clear

While the headlines talk about models capable of automating everything, “the reality I see inside companies is completely different,” Francisco explained. “It’s no longer a technological challenge: with the same AI we had three years ago we could be doing plenty of things that still aren’t being done. We’re still far from seeing companies say they’re selling 30% more while spending 30% less — which is what everyone is after.”

The main reason? The one we already analysed when discussing the BYOAI phenomenon: 68% of skilled employees already use AI, but 77% do so with their own tools, unsupervised. “We’ve moved from the experiment to reality, and reality is something else: information leaking into models, hallucinations, work in silos that improves the worker but not the company.”

Governance doesn’t slow innovation down — it turbocharges it

The most counterintuitive idea of the interview: “It was assumed that governing AI, putting limits on it, worked against innovation. It’s exactly the opposite. Companies with a solid governance model are advancing enormously, because they let their people build their own agents and platforms in a controlled environment — and what works can be scaled.”

And for anyone doubting why limits matter, Francisco shared a real test we ran on air: we gave an autonomous agent —with its own email, Teams access and tools, like one more employee— a dashboard to build, and stripped away all governance. The result? It decided it needed a server, signed itself up, activated the account with the email code and paid for a €20 subscription with the preloaded card. “Look where they can go if we don’t control them. It’s like hiring a person: you train them, tell them what they can do and what information they can access. With autonomous agents you have to do exactly the same.”

The rocket and Mars

On where to start, Francisco was clear — the challenge is no longer outside, but inside. “It’s the company that knows its processes and its goals. AI has given us power never seen before: anyone can get a rocket off the ground. The challenge is steering it to Mars.” That’s why our AI Strategic Dig programme starts by sitting down with each company to draw up a two-year plan, with simple KPIs —productivity, costs, customer satisfaction— and milestones where, note, many activities aren’t technical at all, but about adoption and process redesign.

There was also room for ADA and autonomous agents applied to business: from the insurance executive who creates a new product by conversing with the system, to complete bank reconciliations. With one non-negotiable principle: “capitalise your company’s knowledge in a brain of your own, which is your asset — don’t hand it over to ChatGPT.”

The three decisions to avoid falling behind

At the close, the million-dollar question: what should a CEO who hasn’t taken the step do tomorrow? Francisco’s answer, in three moves. One, a two-year plan — “not one, not three”. Two, an AI governance model, “because it massively incentivises internal innovation”. And three, training: “the moment we lose our fear of AI and learn to steer it, we’ll unlock its full power.”

You can watch the full interview here. And if your company wants to move from pilots to results, you know where to start — let’s talk.