AI literacy is no longer just a nice-to-have. From 2 February 2025, AI literacy will even be a legal requirement. Article 4 of the AI Act obliges companies to ensure their employees are AI literate. But that's not all: organizations and their employees must also develop a sound understanding of artificial intelligence (AI) and - related to this - a wide range of skills in order to remain competitive and successful in the long term.
Our AI literacy canvas offers a methodical framework for the targeted management of initial skills development and continuous AI training within the company. It enables both standardized AI training courses and individual AI courses to be planned precisely and helps to make the right decisions - whether in terms of learning content, teaching formats, or in connection with measuring progress and evaluating results. The canvas can be used at various levels within the organization, from individual employees to teams or departments. Working with the canvas guarantees that people are at the center of AI training so that the AI literacy push is a complete success with a high level of acceptance within the company.
A canvas is an effective and widely used method for structuring problems, visualizing ideas, and guiding creative processes. It enables a clear representation of complex relationships and creates a common understanding within a team or organization. It also makes it easier to clearly identify and derive needs, framework conditions, resources, dependencies, and specific measures.
Originally, the idea of collecting information on a canvas to clarify relationships comes from business model development and product design. Alexander Osterwalder's Business Model Canvas and the Value Proposition Canvas in particular have become widely recognized.
In the following, we will guide you through the nine components of the AI literacy canvas. The structure within the canvas will give you insights into the connections and dependencies that you need to be aware of in order to develop a value-adding AI literacy strategy. We will highlight the connections between the individual fields and provide you with concrete examples to make it easier to understand.
In the top right-hand field of the AI literacy canvas, we start with the first step: identifying the key players. Here you record which people within your organization work directly with the AI applications or will benefit from the results. This can include a wide range of individuals - from IT experts to analysts, but also people from completely different areas of the organization. Clearly defining the users is crucial to effectively address the next steps.
This field should be worked on first so that you can always focus on the needs of the users as you continue to fill in the canvas. Positioning it on the far right ensures that you can always read the completed canvas from left to right, with your users remaining visible as the ultimate goal that everything is working towards.
You can enter personas here that represent the different types of users, as well as specific names of individual employees who will be working directly with the AI applications. You could also enter the name of a department or a department abbreviation to make it clear which teams are affected by the AI initiatives. The exact choice depends on the individual situation and the specific requirements of your organization.
On the far left of the AI literacy canvas, you define the AI use cases, as these represent the starting point for further work with the canvas. The use cases on the far left, together with the users on the far right, form a brace that holds the entire canvas together. Both influence all further steps in the canvas.
Here we concretise how and in which areas the identified users will use AI. The development of use cases in direct cooperation with the users ensures that the AI technology is used in a targeted and effective manner and that the skills that are actually required are acquired.
Some examples of use cases could be: ‘Recommendation engine for our media library’, ‘Automated answering of questions to customer service’, or ‘Dynamic pricing project for seasonal goods’.
In the AI literacy canvas, to the right of the use cases - and as a supplement to them - is the area for everything to do with data and technologies. This is where you enter which AI methods and technologies are to be used. These answers should be closely aligned with the defined use cases to ensure that the methodological and technological competencies that are acquired actually contribute to the intended use cases.
Data sources such as ‘Web Analytics’, ‘CRM’ or ‘Product Information System’ can be mentioned here, as well as tools such as ‘AWS components’, ‘Apache Spark’ or ‘MS Copilot’. In addition, methods such as ‘Collaborative Filter’, ‘Prompting for LLMs’, or ‘Gradient Boosting (XGBoost)’ can be specified.
To the right of ‘Methods, data, tools’ is the area in which the knowledge to be acquired and the required competencies are to be specified. The skills and knowledge that users need to be able to use the aforementioned AI methods and technologies effectively are specified here. The successful development of targeted training measures depends to a large extent on the objectives being clearly defined. This can also be done based on user stories, which are structured according to the following pattern:
I as [PERSON/GROUP]
need the competence [CAPABILITY],
to achieve [GOAL].
‘As a manager, I need basic AI storytelling techniques to get my team excited about projects and to communicate the results of our work to my manager in an understandable way.’ ‘Using ChatGPT and critically evaluating the results’, ’Being able to train a recommendation engine based on our existing data.’
The area for resources is located in the bottom left-hand center of the canvas. This reflects what internal and external resources are already available to support the AI literacy initiatives. This can include budget, personnel, materials, and external partnerships. Existing training tools, formats, or established training processes should also be recorded here.
As an example, you could note here that there is already an existing examination software, licenses for an external training platform are available, a paid account for ChatGPT exists, a 14-part video course called ‘Python with Bernhard and Tim’ can be found on the intranet, a data storytelling coach named Julia works in department DS119 and that the 4-week onboarding phase can be used to carry out basic training.
At the bottom right of the canvas, you will find the area for opportunities and goals. This broadens the perspective and looks at the long-term benefits that can be achieved by increasing AI literacy. The aim here is to determine the extent to which the new knowledge can contribute to the company's strategic goals in order to derive the urgency of the measures and the intensity of the training projects in the form of specific formats in the next step.
You could write down here that the goal of the AI literacy initiative is to remain competitive by increasing internal efficiency, to finally realize a new business model, or to offer employees future prospects and development opportunities.
In the upper right-hand center of the canvas is the area for specific education and training formats in which the training is ultimately to take place. This is where the theoretical and sometimes still vague concerns and objectives are translated into realizable practice. The selection of formats should be flexible and orientated towards the needs of the users.
Here you could enter that a weekly evening course on the basics of AI (with practical examples and room for experimentation) is offered - which lasts a total of 3 months -, a basic Python course is available on demand from external providers, a ‘Digital Ethics’ certificate course can be completed at the HWZ Zurich or a pilot project on peer-2-peer learning is being carried out.
Directly below the formats is the assessment and progress section, which is closely linked to the learning formats, as the assessment of learning success often depends directly on the chosen teaching methods. In this section, mechanisms can be developed and collected to measure and assess progress in relation to AI literacy. KPIs or specific targets are set and a review schedule is established. This allows the strategy to be flexibly adapted and the success of the entire AI literacy initiative to be measured.
The following could serve as practical examples: Doubling the number of certifications for the AI tool ‘TrainCopterPlane’ by the third quarter, filling the role of ‘Person responsible for data protection in AI applications’ and appointing two deputies, ensuring equal representation on the ‘AI task force’ with five developers and five analysts by September, ensuring that one person in the department has basic Python skills by the end of the year or that all department members should have successfully completed the basic AI course by the end of the second quarter.
The bottom left-hand section of the canvas is intended for foreseeable costs and risks. This is where you assess the financial and time expenditure as well as the potential risks and dependencies that may arise in connection with the AI literacy initiative. If you identify the obstacles and potential risks during the development of the AI literacy strategy and record them in writing, critical scenarios can be recognized and addressed at an early stage - danger recognized, danger averted!
For example, it could be that the use of AI in human resources harbors risks, an external Python course for the entire department costs €5,000, the quality of the CRM data may be too poor for an in-house AI application, some employees may not be able to take part in further training for fear of the exam and how to counteract this, as well as the opinion of the works council on the planned projects.
If you go through each step carefully and in the order given, you can use this canvas to create a solid foundation for developing AI literacy in your organization.
When working with a canvas, it should first be clearly defined which problem or question is to be solved with the help of the canvas. A clear objective is crucial in order to maintain focus during the process. In addition, all relevant stakeholders should be involved in the process to ensure a multifaceted perspective.
Information that is relevant to the issue can be collected and structured in advance, especially if the issue is a large one. To this end, a responsible person should evaluate which topics could become important in the canvas and initiate corresponding collections of material - via questionnaires, interviews, or, in the case of larger projects, even in separate workshops for the corresponding fields in the canvas.
The various areas of the canvas can then be filled in directly with a felt-tip pen on a large printed canvas, with the help of post-its or in digital form (e.g. in Miro or Mural) and linked together in a specially planned workshop.
Openness to new ideas and unconventional approaches is a key success factor when working with a canvas. The content created can be validated through feedback from other team members or through tests with potential users. Based on this feedback, adjustments can (and should!) be made iteratively after the workshop and over the coming weeks and months.
Did you just read ‘iteratively’? Correct. You probably know the word from the context of agile methods.
The user-centered design of this canvas means that all relevant aspects of AI literacy development are taken into account and presented on a visual level. This structure promotes dynamic collaboration within the team and enables rapid adaptation to changing requirements and insights. By regularly updating the AI literacy canvas, organizations can ensure that their AI training strategy is continuously improved to meet the needs of users and promote innovation. In a fast-moving and complex technology landscape, this agility is crucial to remain competitive and ensure long-term success.
You can work with this canvas as part of a workshop with all stakeholders involved and use it to visualize the results on a permanent basis. It should remain a ‘living document’ that is regularly updated and adapted to the current tool and technology landscape of your company. This is the only way to ensure that it does not become outdated and that an AI training strategy can be aligned with it in the long term.
Overall, the successful use of a canvas is an iterative process that requires a clear objective, a structured approach, and the involvement of all relevant stakeholders. By taking a systematic approach and being prepared to accept feedback and adapt the canvas accordingly, valuable insights can be gained and well-founded decisions made.
Have questions? Need some support getting started with our AI literacy canvas?
Check out our AI literacy training, or get in touch and let's find out together how we can help!