Cisco's position in AI education is unusual and it took me a while to work out what they are actually doing. They are not trying to teach you to build models, and they would lose that fight to DeepLearning.AI and Hugging Face if they tried. What they have instead is the physical layer, the networks and data centres that AI workloads run on, and a training arm with decades of reach into schools, colleges and career changers. The recent AI push splits along exactly that line, with free literacy courses at the front door and infrastructure focused certifications at the back.
Introduction to Modern AI is the free course most people will meet first, and it is better than it needs to be. It runs a few hours, requires no programming, and covers the basics of AI, machine learning and models before putting you in front of a range of actual tools, including chatbots from several different vendors rather than just one. That last detail is worth calling out. Most free vendor courses use their own product as the only example, and learners come away thinking the product is the technology.
Cisco having you compare outputs across different assistants teaches something more useful, which is that these things differ and that comparing them is a skill. It is available in a lot of languages, it is free, and it issues a badge. For someone who has never touched this subject, a few hours here is well spent. It is also, and this is the honest part, shallow.
You will finish it and be ready for something harder within a week. Treat it as an on ramp and have your next course chosen before you start, or it will end up as another badge in a collection that goes nowhere. The certification side is newer and more interesting. Cisco has introduced an AI practitioner level certification with an exam at roughly $150 and a sixty minute format, covering generative AI models, prompting, ethics and security, and agentic AI.
It is valid for three years like most of their certifications. There is also a data centre AI infrastructure specialist track, and that one is where I think Cisco is genuinely differentiated. Almost nobody is teaching the networking, power, cooling and fabric side of AI workloads in a structured way, and the people who understand it are in short supply and paid accordingly. If you already work in data centre or network engineering, that path is a much better use of your time than another generic prompt engineering course.
The practitioner certification is harder to be enthusiastic about. The syllabus reads like a competent summary of material available free from a dozen sources, and it is new enough that no employer has yet formed a view on whether holding it means anything. Certifications acquire value slowly, through hiring managers seeing them repeatedly on CVs of people who turned out to be good, and this one has not had time to do that. At $150 it is a small bet rather than a big one, but do not expect it to open doors in the way a CCNA does in networking, at least not yet.
My last complaint is administrative and it is the same one I have about IBM SkillsBuild. Cisco has Networking Academy, Cisco U, the Learning Network and the certification site, some courses are co branded with IBM, and working out which platform holds the thing you want is genuinely tedious. Start at netacad.com for the free courses and at Cisco's certification pages for the exams, and expect to be bounced between logins. The learning is good, the plumbing is not.