Executive AI program or one-to-one: start with what you want to do on Monday

Two questions hide inside this choice. One is whether you need a view on AI you can defend in a board meeting. The other is whether you need to get something out of your own files by Friday. Executive programs at business schools are built for the first question. Below is a five-part test that sorts which one you are actually asking.

Five tests that sort the two questions apart

Start with one sentence: what you want to be able to do at your desk next week. A sentence about holding a view in front of a board points one way. A sentence that names a document you build by hand every quarter points the other way.

Then run these five tests on any offer you are considering, and on whatever a peer recommends next month. Each one is specific enough that a seller can answer yes, no, or not covered.

• A document from your own company enters the room, and at least one working hour is spent inside it.
• Someone watches you judge an output from your own file and corrects your judgment, not your typing.
• The data line gets drawn for your company, against your own contracts and your group policy, not in general.
• You can ask a basic question and get a straight answer, with nobody who reports to you in the room.
• When it ends, something of yours is still running, rather than sitting in a folder of notes and slides.

Four of the five are about your own material. That is deliberate. Strategy content on AI is widely available now and much of it is good. Unhurried practice on your own files is the scarce part.

What does an executive AI program give you?

A serious grounding. The programs run by Indian business schools range from a few months to most of a year, delivered live-online, self-paced or blended, sometimes with a few days on campus. They mix recorded content, live sessions with faculty and industry experts, and demos. Most close with a project or capstone on a real business problem in your domain, and some offer a specialization track by function.

Read that against the five tests. The fifth test depends on what you take into the closing project, because a project is an asset only if it runs on work you will still be doing in March. The first test is the one the format strains against. Months of content, live sessions and demos give you a real grounding in what the technology is and where it pays. None of it puts your own quarterly review file on the screen for an hour.

That is not a criticism. A curriculum written for professionals across functions cannot be written around your functions. It is what the purchase is for.

How do you choose between two programs?

On calendar and specialization rather than quality. If a program offers a track that matches the function you already run, and you can block the campus days, that shape fits you. If you want breadth across tools and a format you can take around a travel calendar, a live-online and self-paced program fits you better. Either puts a credible name on your profile and gives you peers at your own level to argue with, which is often the real value of an executive program.

What no cohort format can do is stop for an hour inside one participant's own review file. A capstone chosen from a track sits close to your work without being your work. The data question has no general answer either. Yours depends on your contracts, your group policy and your legal position, and no syllabus written for every sector can land on it.

Where the hour on your own file sits

The one-to-one shape is the exception on the first and third tests, and it is an exception with real costs. No cohort, so no peer group. No institutional name on the certificate. No breadth, so you will not leave with a survey of agentic architectures.

What you get instead is the hour. Take research work a senior leader actually carries: a comparison table of the 10 to 15 leading players in a market your company is weighing, with an executive summary and source URLs, then three separate research reports combined into one master table with the contradictions resolved. In an A.D.A.P.T. AI Coaching session built around you, the sector is yours, the market is the one your management committee asked about last month, and the checks are yours to run. You read the master table against the source pages before the summary exists. You find the row that is wrong, because there usually is one, and then you fix the table rather than explain the summary away.

That is the hour a cohort cannot give you. The tests above are still the useful part. Run them on any program, on the one-to-one shape, and on whatever comes next.

What else do senior leaders ask before picking one?

Is a six-month program realistic with my travel?

Most executive formats are built for people who travel, with self-paced content between live sessions, so the question is not the total length. It is whether you will keep the weekly habit in a bad quarter. Check how much of the content is fixed-date and plan around those dates only.

Will a program certificate answer the question my board asks?

A business-school certificate answers where your view came from, and that is worth having when the board asks. It says nothing about whether you can catch a confidently wrong output about your own numbers. Work out which of the two you are actually short of, then buy that one.

Can I start on something before the next cohort opens?

Yes, and it costs nothing. Take the report you read every Monday, ask ChatGPT for the three items in it that need a decision from you, then check all three against the report yourself. If one of the three is wrong and you caught it, you have started building the evaluation muscle.