I just cannot bring myself to write another opening paragraph about how AI is everywhere; how transformative it is; and how all things education will go down like the Titanic unless we innovate faster than lightning and engage in a fierce head-to-head battle with digital-only providers.
All of those things may very well be true, but regurgitating the same old truisms will almost certainly get us nowhere. On the other hand, doing nothing will get us nowhere, too.
In March, I wrote a post titled Develop Your AI Strategy for Higher Education. I still think that post is worth consulting, but a lot has happened since March.
Many students now utilize powerful AI models in their daily lives for all sorts of purposes, and they use AI and chatbots for educational activities. This fact alone means that now is the time to think concretely about AI in education if you haven’t done so before.
Today I want to direct your attention to seven areas of AI you probably need to consider very soon. Along with each area you will find a few actionable recommendations for you and your team to get going.
Here goes.
1. Build a Shared Understanding of AI
It is crucial to establish a common understanding of AI among all stakeholders in your institution: teachers, admin staff, leadership, etc. AI encompasses many complex technologies, so agreeing on possibilities and challenges in your specific educational context is vital.
Actions
Host inspirational talks, workshops, facilitate teaching experiments with AI, and consider long-term expert partnerships to build understanding.
Identify colleagues already working with AI and learn from their experiences.
Discuss informative videos, blogs, etc. to inspire each other.
Create a shared overview of AI tools, evaluating them on parameters like pedagogical potential, accessibility, UX, security, etc. Start small, i.e. with a shared Excel sheet.
2. New Approaches to Pedagogical Use of AI
AI can help develop general skills like critical thinking, while also enhancing subject-specific learning. AI should never limit educational outcomes, but it can improve, augment and complement human skills in teaching. Evaluating credibility and relevance will be increasingly important for students.
Actions
Appoint pedagogical/didactic AI leads in teaching teams.
Think about integrating AI capabilities into existing subjects rather than developing entirely new AI subjects.
Teachers should jointly decide where AI can and cannot be incorporated into curricula.
Continuously evaluate AI tools against learning goals and adjust usage accordingly. AI should enrich, not replace, traditional teaching.
3. Reconsider Examination Forms
Education has long emphasized written exams, prioritizing output over processes, as I wrote in the recent post Output is Dead, Long Live Process-Oriented Learning! But skills like creativity and critical thinking are difficult to quantify. A focus on output often overlooks the importance of creative problem-solving, experimentation and collaboration.
Actions
Do not wait for accurate AI plagiarism screening tools. These are largely ineffective and prone to false accusations. Check out
‘s excellent post on the topic here.Consider more process-oriented assessments like portfolios or oral exams focusing on critical dialogue, argumentation and problem solving.
Carefully examine whether written exams truly are the best assessment method for each subject.
4. Equal Access for All
All students must have equal access to AI technologies, including those with disabilities, learning impairments, etc. This makes AI an institutional issue requiring guidelines and teacher training on supporting diverse students using AI responsibly. Teachers must also be aware of biases, falsehoods generated by AI ("hallucinations"), and lack of diversity in AI systems.
Actions
Involve students and staff in identifying discrimination or inequality in educational AI contexts.
Critically review all AI-generated content for biases before classroom use.
Teach students to identify and challenge AI biases.
Have open discussions on AI consequences. No one has all the answers; open dialogue is key.
5. Ethical Considerations
AI systems represent a particular worldview, trained primarily on white, Western, male-authored texts. This leads to issues like gender, racial, and other biases when AI responds to societal questions. These ideological biases inherent in AI systems should be addressed whenever the technology is used.
Actions
Make students and teachers aware of AI systems' ideological biases.
Train students across disciplines to identify potential biases in AI content.
Develop clear institutional AI guidelines before implementation, with input from all stakeholders.
Use providers who actively address ethical concerns with their AI products/services.
6. Data and Consent
AI systems' complexity necessitates a legal, data-responsible approach to their educational use. Not all platforms are secure or appropriate. Institutions must enable and constrain student and staff use of AI technologies responsibly.
Actions
Ensure generative AI use involves informed opt-in consent. While this undoubtedly complicates processes, consent is vital given AI's ethical complexity.
Provide informed consent information, potentially as part of an institutional AI policy, with consent retraction options.
Maintain a register of AI tools used by teachers. This could well be identical to the overview recommended in 1).
7. Who Should Own AI in the Organization?
AI cannot be managed like other IT systems — its ubiquity makes end users primarily responsible for its use. Institutions should provide AI guide and ensure compliance, but it is key that individuals use it responsibly. Empowering teachers with AI ownership and co-responsibility will make it a natural part of pedagogical practice.
Actions
Anchor AI broadly within teaching and with individual teachers. Teachers will need to adapt to a role of AI consultants whether they like it or not.
Provide technical support so teachers are not isolated when using AI tools.
Consider AI mentors to assist colleagues as superusers.
Make clear that AI responsibility lies with teachers themselves, not IT departments.
Conclusion
Hopefully, these recommendations for higher education will get you on the right track for adopting AI in your institution. AI poses complex challenges, and this guidance will only steer you in the right direction. But it is my hope that it will do just that.
As always, reach out to me with any questions or comments you may have.
Thanks for reading along!