You are behind on AI if your company has no working strategy, your employees are untrained or using AI in secret, and you have no plan for the EU AI Act. Everything else that looks like being behind is marketing noise. And if you are genuinely behind, that may be the strongest position you have held in years.
The signals that do not count
Most leaders judge whether they are behind by watching what other companies announce. By that measure, being ahead looks like this: the word AI added to every product and platform name, an AI strategy announced in press releases and social campaigns, and AI partnerships declared with clients and existing partners.
None of it counts. Putting AI into an existing brand name does not make it AI. Announcing a partnership built on a thin AI feature is not a strategy. It is noise, and the loudest announcements often signal the weakest strategies. If the only thing separating you from your competitors is that you have not performed this theatre, you have not fallen behind. You have skipped the noise. And if a competitor has just declared themselves AI-first, read that announcement as intelligence rather than as a threat.
The signals that do count
Three honest tests tell you where you actually stand. First, does a working AI strategy exist, owned by leadership, connected to how the company makes money? Second, are your employees trained and using AI openly, or are they using it in secret on personal accounts while the official position sits in a slide deck? Third, do you have a compliance position on the EU AI Act, or has nobody read it?
Those three tests measure the company. A harder question sits underneath them, and it is personal: whether AI is about to expose the leadership rather than replace it.
If you fail all three, you are behind. Now let me tell you why that may be good news.
I know the cost of being early. I paid it.
In 2017 I founded the IBM Watson AI Lab for Cancer at the Oslo Cancer Cluster, one of the first AI-in-cancer-research institutions on a European cancer campus. It was early, and being early carried a cost that nobody warns pioneers about. We had to educate a market that did not yet have the words for what we were doing. New terminology, new ways of thinking, new ways of helping people, all of it explained from zero, year after year, at real expense. Only now is the market catching up to what that lab started.
Pioneers earn the credit. They also carry the heaviest lift. Most companies should not want to be first, and the ones agonising about being last have usually never counted what first actually costs.
You are late to a party that just got good
Think of the best party you have ever been to. It rarely got interesting before midnight. Arriving late did not mean you missed it. It meant you walked in at the right time.
AI in the enterprise has been that party. For the past few years it ran on models that were good at best, prone to mistakes, and nowhere near the reliability the marketing claimed. Companies that rushed in early paid for that gap: budgets released without direction, experimental projects, token bills, implementation fees, and consultants learning on the client’s money. Many of those companies are now cutting their AI budgets, which tells you how those early bets performed.
Arriving now, you inherit three advantages they never had. The failures are documented, so you can learn what not to do from other companies’ expensive lessons. The frontier models are finally, in my view, worthy of enterprise work. And you can build your strategy once, on solid ground, instead of rebuilding it every time the technology lurches forward.
I have sat with companies that were genuinely two years behind, and watched them overtake the ones who moved first. Not because they were cleverer, but because they built once, on ground the early movers had already tested and paid for.
The advantage has an expiry date
Here is where I must be honest with you, because this is where most reassurance turns into a trap. Lateness is only an advantage if you move now. Late and slow is fatal. The competitors ahead of you are refining what they built, and organisational learning compounds in a way you cannot buy later.
And there is a second clock running that has nothing to do with competition. The EU AI Act is law. Its transparency obligations apply from 2 August 2026, and while the high-risk deadlines have moved to late 2027 and 2028, the penalties have not softened: up to EUR 35 million or 7 per cent of global turnover. The obligations follow what your AI does, not how large your company is. An AI strategy is no longer something you build because your competitors have one. It is something you build because doing it wrong is now a regulated, finable event.
So the position is this. You did not pay the pioneer’s bill. You did not burn the experimental budgets. You arrive as the models come good and the rulebook becomes clear. Before any of it, do the honest internal reckoning first, then build the strategy, train your people properly, and move. Your lateness is a gift. It stays a gift only for as long as you treat it like one.
Frequently asked questions
How do I know if my company is behind on AI?
Ignore the marketing signals. You are behind if you have no working AI strategy owned by leadership, your employees are untrained or using AI only in secret, and you have no plan for the EU AI Act. AI in product names and press releases tells you nothing.
Is it too late to start with AI in 2026?
No. Companies starting now skip the expensive experimental years, learn from other companies’ documented failures, and deploy on frontier models that are finally strong enough for enterprise work. The advantage disappears only if you keep waiting.
What does it cost to be too early to AI?
Years of educating a market that lacks the vocabulary for what you are building, and heavy spending before the technology matures. I lived this founding the IBM Watson AI Lab for Cancer in 2017. Pioneers earn the credit, and they pay for it.
Can my company be fined under the EU AI Act?
Yes. The Act is law, transparency obligations apply from 2 August 2026, and penalties reach EUR 35 million or 7 per cent of global turnover. The delayed deadlines for high-risk systems do not remove the liability.
Thomas Anglero is a Strategic AI Advisor, keynote speaker and author of Intro to Artificial Intelligence. He has delivered over 450 keynotes across 30 countries for organisations including IBM, the WHO, the World Government Summit and the European Commission. He founded the IBM Watson AI Lab for Cancer at the Oslo Cancer Cluster and closed over $500 million in enterprise transformation deals as CTO and Chief Innovation Officer at Cognizant.
If you are leading your organisation through this, I work with a limited number of senior leaders each quarter. Get in touch at Anglero.com.