AWS's new AI cert tests judgment, not code

AWS just put a certification for people who do not write code in its catalog. AWS Certified AI Business Strategist (AIB-C01) opened for beta registration September 1 and starts delivering September 29. No coding required, no AWS implementation experience, no prerequisite certs. The exam does not test AWS services. It tests whether you can decide which AI work is worth doing and govern it once it ships. Coming from the company whose entire certification business is built on proving you can configure things, that is a real signal about where AI projects fail.

What actually happened

Per the AWS Training and Certification blog post dated September 1, the exam targets line-of-business leaders, sales professionals, consultants, and program managers — people who align AI investments to outcomes rather than deploy them.

The mechanics from the certification page:

• 85 questions, multiple choice and multiple response, 170 minutes
• $50 during beta, $100 standard
• Six months of experience working with or alongside AI initiatives is the recommendation; there is no coding or AWS prerequisite
• Four domains: AI fundamentals and literacy, AI strategy and business value creation (the largest), AI governance and responsible AI leadership, and business readiness and AI transformation
• Certify by February 15, 2027 and you get an Early Adopter badge on top of the standard one

AWS is explicit that its own services show up as context, not as the subject. The stated aim is judgment that is portable across organizations.

Why it matters for your business

We have watched enough AI pilots die to say this plainly: almost none of them failed on the model. They failed because nobody could say what the thing was supposed to improve, nobody owned the decision when the output was wrong, and nobody had a number to compare against six months later. Those are the four domains on this exam. AWS did not invent that insight — it read the same failure reports everyone else did and priced a credential against it.

Two honest reads, in order.

The signal is worth more than the badge. A 170-minute multiple-choice exam does not prove someone can run an AI program, and hiring on it would be a mistake. What it does tell you is that the largest cloud vendor now believes the binding constraint on AI adoption is business judgment, not implementation skill. That matches what McKinsey found when a third of surveyed organizations skipped a software purchase because they could build it with coding agents — the build got cheap, so the deciding got expensive.

The exam outline is free and more useful than the certificate. Read the four domains, then audit your own last AI project against them. Did you write a business case with a number in it? Who signs off when an agent produces a wrong answer for a customer? What is your rollback? If those questions have no owner, a $100 exam will not create one — but the outline is a decent checklist, and the checklist is the part you actually need.

For a small team, the honest advice is: skip the cert, steal the syllabus. Then spend the $100 on a scoped pilot with a metric attached, which teaches the same thing and leaves you with a working system instead of a badge.

Sources: AWS Training and Certification blog, AWS Certified AI Business Strategist exam page.