SkillsBuild started as IBM's corporate social responsibility programme, which explains both its strengths and the odd shape of it. The material is free because IBM wants the goodwill and the talent pipeline, not because it is a stripped back version of something they sell you later, and that changes the tone. There is no urgency, no countdown timer, no scarcity language, and nobody is going to email you about a limited time offer. After spending a lot of time on paid platforms, the absence of that pressure is genuinely pleasant.
The AI Fundamentals path is the piece most people arrive for and it is the piece I would recommend. It runs through what machine learning actually is, how neural networks fit into that picture, natural language processing, computer vision, chatbots, and a decent section on AI ethics and bias that is better than the throwaway ethics module most courses bolt on at the end. There is no Python, no maths beyond arithmetic, and no environment to configure. That is a design decision and it is the right one for the audience, which is people who need to understand AI well enough to have an opinion in a meeting, not people who need to train a model.
Around ten to twenty hours is a fair estimate for most learners, and you can spread it across a couple of weekends without losing the thread. Where I would push back is on how this credential gets talked about. There is a genre of LinkedIn post treating an IBM SkillsBuild badge as a career milestone, and a genre of blog post that follows it, and both are wrong in the same direction. The assessments are multiple choice, they are forgiving, and you can retake them.
Nothing you do is verified by a human or checked against work you produced. That makes the badge a signal that you spent some hours on the topic, which is a real if modest signal, and not a signal that you can do anything in particular. Hiring managers in technical roles know this. Hiring managers in non technical roles often do not, which is exactly why it can still be useful on a CV if you are moving into an operations, marketing or project management role where AI literacy is now expected.
The other thing worth knowing before you start is that SkillsBuild is not one platform, it is a front door to several. Some courses live on skillsbuild.org, some are delivered through the Cisco Networking Academy platform, badges are issued through Credly, and IBM's separate professional training site uses similar naming for different products. Working out which login goes with which course is more annoying than any of the actual learning, and I lost a genuinely irritating amount of time to it. If you hit a dead link or a course that will not open, it is usually because you are on the wrong one of IBM's several education properties rather than because anything is broken.
My honest position is that this sits in the same bracket as Google AI Essentials and Elements of AI. It is a good free introduction that respects your time, it will make you meaningfully less confused about what these systems do, and it is not a stepping stone to building anything. If you want the next step after this, the answer is Python and a project, not another badge. If you are a student or job seeker who wants some evidence of initiative that cost nothing but a couple of weekends, it does that job perfectly well, and IBM deserve some credit for building it and then not charging for it.