--- title: "Chief AI Officer: the role doesn't belong in IT | Carl-Uwe Würfel" description: "A Chief AI Officer at a mid-market company needs the authority to change roles and workflows, or the job is software admin. Scope, limits and four checks." url: https://carl-wuerfel.de/en/chief-ai-officer author: Carl-Uwe Würfel --- 1. [Home](https://carl-wuerfel.de/en) 2. [Knowledge](https://carl-wuerfel.de/en/knowledge) 3. Chief AI Officer # Chief AI Officer in the middle market: whoever runs the workflows runs the AI. A Chief AI Officer is judged by whether work across the company changes, so the role belongs on the leadership team and not in IT. The obvious move is a new hire who picks tools and runs pilots. You get new software and the work stays the same, because decisions about roles and workflows get made somewhere else. At SUMAX I have held the role since 2023 on top of my job as COO, and the large majority of our employees now use our own AI tools every month. [Carl-Uwe Würfel](https://carl-wuerfel.de/en/about)Updated November 7, 2026[Deutsche Fassung](https://carl-wuerfel.de/chief-ai-officer) ## On this page 1. Without authority over roles, a Chief AI Officer just manages software. 2. The first tool comes after the foundation. 3. In the middle market, the role fits the COO. 4. An AI officer is no substitute for a Chief AI Officer. 5. A tool nobody opens doesn't count. 6. Four questions show whether the role is set up right. 7. Frequently asked questions > Whoever passes on an AI result has checked it and stands behind it. Enforcing that rule is leadership work, not technology. ## Without authority over roles, a Chief AI Officer just manages software. **Takeaway** Put the role on the leadership team, with the authority to change roles and workflows. The role has four parts: one shared platform instead of scattered tools, rules for approval and data access, the cost of every AI call, and redesigning roles and workflows. Platform, rules and cost can be delegated. The decision on roles cannot. A team can draft the new role profiles, but decisions on organization and people stay with leadership. Without that fourth part, the change stops at the tools. There is new software and everyone works the way they did before. In a mid-market company, a Chief AI Officer without a seat on the leadership team is half a solution. Platform, rules and cost can be delegated, the decision on roles cannot. _Figure: The Chief AI Officer on the leadership team owns four parts: platform, rules, cost, and roles and workflows. The first three can be delegated. When roles are redesigned, the drafting can be delegated, while decisions on organization and people stay with leadership. Without that fourth part, the change stops at the tools._ ## The first tool comes after the foundation. **Takeaway** Settle sign-in, permissions and access for everyone before any department gets a tool of its own. The work starts with the foundation for everyone: who can sign in, who sees which data and which access applies to which tool. Security is part of it from day one. At SUMAX a layer of our own sits between the applications and the AI models. It lets us switch models, and it logs every call with its cost. Tools for single tasks came after that. Process knowledge that used to be spread across heads and spreadsheets now lives in one place. The tools are built with the people who use them, because they know best which step eats up their time. ## In the middle market, the role fits the COO. **Takeaway** Whoever owns the workflows should own the AI inside them. The change hits workflows, roles and training first, which is the operating business. Whoever owns the workflows sees where AI can take work off people's plates and can adjust roles directly. That is why I took on the role at SUMAX in 2023, on top of my job as COO. I built the technical foundation with the team. We rewrote the role descriptions, and training uses real tasks from each team's own work. AI takes on routine work, and the people stay. Justify the role with headcount savings and you lose the people who have to check the AI's output. A separate position pays off more when many workflows change at the same time. ## An AI officer is no substitute for a Chief AI Officer. **Takeaway** Training can sit with an AI officer. The redesign sits with leadership. Since February 2, 2025, Article 4 of the EU [AI Act](https://carl-wuerfel.de/en/ai-act) has required providers and deployers to take measures for their staff's AI literacy. The regulation does not require an AI officer. An AI officer can run training and maintain a policy. That helps, but it doesn't change a single job description. A Chief Data Officer looks after [data quality](https://carl-wuerfel.de/en/glossary/datenqualitaet) and data management, and a Head of AI leads a team and the technical build. The Chief AI Officer answers, at leadership level, for whether work across the company changes. Mix these roles up and you get training and software, but no redesign. ## A tool nobody opens doesn't count. **Takeaway** Measure the role by usage, cost per call and output per working hour. The yardstick is everyday use. At SUMAX most employees turn to our own AI tools every month. A five-figure number of AI calls a month runs through the same layer, each logged with its cost. We replaced several purchased tools with our own, and each editorial hour now produces a multiple of the expert texts it did in 2024. A tool nobody opens still costs upkeep and licenses, so the number of tools rolled out says little. What matters is who uses them regularly and which task gets done faster because of them. ## Four questions show whether the role is set up right. **Takeaway** Go through the four questions at your next leadership meeting and start with the first no. A checklist for owners and leadership teams. One: can the person accountable for AI change job descriptions, or do they have to ask? Two: does every AI request run through a shared foundation with permissions and cost tracking, or does each department have its own accounts? Three: does anyone know how many employees use the tools each month? Four: do the people who pass on AI output have time set aside to check it? Every no marks a spot where the change gets stuck at the tools. I'd rank the first question highest, because without it nobody enforces the other three. ## From practice. * [Evidence**The transformation at SUMAX** Four layers, the flow of AI calls and the model switch.](https://carl-wuerfel.de/en/transformation) * [Person**About** Career, roles since 2018 and 2023, publications.](https://carl-wuerfel.de/en/about) ## Frequently asked questions What does a Chief AI Officer do? A Chief AI Officer is accountable for AI working across the whole company, and above all that means roles and workflows change. The job also covers a shared platform, rules for approval and data, and the cost of every AI call. It starts with the foundation for everyone, meaning sign-in, permissions and access, and tools for single tasks come after that. Does a mid-market company need a Chief AI Officer? It needs someone on the leadership team who owns the change, and that does not have to be a separate hire. A COO or another executive can take on the role; at SUMAX I have held it since 2023 alongside my job as COO. Without that ownership, individual departments build their own tools with no shared rules, access or view of cost. How is a Chief AI Officer different from an AI officer and a Chief Data Officer? An AI officer runs training and keeps the rules current, for example for AI literacy under Article 4 of the EU AI Act, and the role is not mandatory. A Chief Data Officer owns data quality and data management. The Chief AI Officer is accountable for workflows and roles changing, and decides on platform and budget to make that happen. What qualifications does a Chief AI Officer need? They have to know how the company runs and have the authority to change roles. Add leadership experience and enough technical understanding to judge what an AI model can do and where it gets things wrong. Coding is not required, and there is no set career path. How does a Chief AI Officer work across a group of companies? By putting one shared foundation in place for all companies: sign-in, permissions, access and a layer that logs every AI call with its cost. On top of it, each company gets tools for its own workflows. Cost and usage stay comparable, and an acquisition plugs into the same foundation instead of bringing another island. Who does a Chief AI Officer report to? In a mid-market company the role sits on the leadership team or reports directly to the CEO and the owners. That is where decisions on workflows, roles and cost get made anyway. At SUMAX the role is part of the leadership team. ## Read on * [**The transformation at SUMAX**](https://carl-wuerfel.de/en/transformation) * [**Build vs. buy AI: only the build got cheaper**](https://carl-wuerfel.de/en/notes/buy-or-build-ai) * [**The EU AI Act for mid-sized companies**](https://carl-wuerfel.de/en/ai-act) ## Sources 1. [Regulation (EU) 2024/1689 (AI Act), Article 4: AI literacy](https://eur-lex.europa.eu/eli/reg/2024/1689/oj) 2. [European Commission: AI literacy questions and answers](https://digital-strategy.ec.europa.eu/en/faqs/ai-literacy-questions-answers) 3. [Bundesnetzagentur (German regulator): AI literacy under Article 4 (German)](https://www.bundesnetzagentur.de/DE/Fachthemen/Digitales/KI/7_Kompetenz/artikel.html)