Executive summary. Artificial intelligence is pushing higher education beyond experimentation into systemic transformation. The challenge is no longer adoption, but how to implement AI responsibly across the institution. Drawing on insights from ESCP’s AI in Higher Education Summit, this article introduces the ABC Framework: Accountable (governance-first), Bold (institution-wide capability building), and Creative (redesigning teaching and academic work around human judgment). As employers demand AI-ready graduates and industries change quickly, universities must move from fragmented initiatives to coherent strategies. AI is not just a tool, but a sociotechnical shift requiring alignment across pedagogy, operations, and leadership. The institutions that succeed will balance accountability, boldness, and creativity.

Key takeaways

• AI in higher education is shifting from tool adoption to institutional design.
• Accountable: Governance must precede deployment to ensure trust and compliance.
• Bold: Capability-building requires system-wide, not pilot-based, approaches.
• Creative: Teaching and academic work must shift from outputs to process, reasoning, and judgment.
• Institutions that align all three pillars will be best positioned for long-term relevance.

---

From AI Awareness to AI Reality

When Sam Altman published his tweet on November 30, 2022, announcing the launch of ChatGPT, our world changed. While AI had long been embedded in everyday systems, generative AI placed unprecedented capabilities directly into the hands of individuals.

In higher education, this shift has been immediate and irreversible. According to a recent study by CarringtonCrisp, 77% of companies expect new graduates to have AI experience, and 58% believe universities are not doing enough to prepare graduates for an AI-enabled workforce.

And these expectations are not theoretical but reflect changes already underway across industries.

At mobility company Sixt, for example, a significant share of software code is now AI-generated, fundamentally changing the role of engineers, from writing code to designing user experiences and making higher-level decisions. Similarly, in the pharmaceutical sector at Sanofi, AI has reduced regulatory submission timelines from 19 weeks to as little as 5, accelerating innovation cycles while still requiring human validation at every stage.

These examples illustrate a broader shift: AI is not just improving productivity. It is redefining the nature of work itself. For universities, this creates a clear necessary challenge: to prepare students for an AI-driven world.

But as with all new technologies, there are trade-offs. While AI has already clearly shown its potential to improve productivity for faculty, increase accessibility and individualise feedback, as well as support research workflows, the risks remain existential. According to the Digital Education Council’s 2025 Global AI Faculty survey, 83% of faculty are concerned that students cannot evaluate AI outputs.

An over-reliance on AI without proper protocol could undermine assessment integrity, reduce critical engagement, and introduce new forms of inequality linked to access and digital literacy. Some studies already show that AI can reduce our critical thinking skills.

The question is therefore not whether to adopt AI, but how to implement it responsibly.

From Experimentation to Institutional Design

Over the past two years at ESCP Business School and the ESCPTech Institute, we have approached this challenge by treating our institution as a living laboratory.

We began with a proof-of-concept phase with OpenAI, providing access to ChatGPT Edu to 1,000 AI champions, including students, faculty and staff. This was later scaled to the entire ESCP community, alongside the integration of the open-source platform Hugging Face, the world’s first higher-education partnership with the platform, enabling not just experimentation, but the ability to build, audit and understand AI systems.

After this period of testing and learning inside the institution, more questions remained. AI is not a topic that can be tackled alone. To move forward, we needed a broader, collective perspective. Therefore, we organised the first AI in Higher Education Summit, bringing together 183 participants, representing 67 universities and more than 27 countries, to explore one central question: how should higher education evolve in an AI-driven world? Over two days, experts across higher education exchanged and debated this question, with 50 contributions published in the SSRN proceedings.

Insights from the AI in Higher Education Summit suggest that the debate is evolving. The focus is shifting from tools to systems, from experimentation to institutional design.

This transition highlights a persistent ‘governance gap’. According to the 2026 Stanford AI Index, only 48% of institutions have policies governing acceptable use of generative AI. And even if policies and guidelines do exist, they are often disconnected from day-to-day practices, leaving faculty uncertain about expectations and responsibilities. At the same time, pedagogical transformation tends to lag behind technological deployment.

According to CarringtonCrisp, 44% of faculty believe students know more about AI than they do. 62% say AI improves teaching, yet 43% worry it is ‘dumbing down’ education.

While comparisons have been made to previous technological advancements in pedagogy, it is clear that AI moves us beyond a technological upgrade. AI should therefore be understood as a sociotechnical transformation affecting governance, pedagogy, capabilities and infrastructure simultaneously. Addressing this complexity requires a coherent, institution-wide response.

Introducing the ABC Framework

To support this transition, the ABC Framework for AI-Ready Universities provides a structured approach to implementation. Alara Tascioglu developed the framework through inductive thematic analysis of the summit contributions, consistent with ESCP’s mission and presented in full in an Impact Paper to be published later this year. We apply it jointly here to synthesise and introduce the proceedings. It is built around three complementary pillars:

• Accountable: AI as a governed sociotechnical system
• Bold: AI as capability expansion through disciplined experimentation
• Creative: redesigning teaching and academic work around human judgment

The core premise is simple: each pillar is necessary, but none is sufficient on its own. The framework is not prescriptive, but offers a way to align institutional strategy, practice, and educational purpose.

Accountable: Governance before tools

AI implementation begins with governance. Too often, institutions begin with tools and attempt to define rules afterwards. The ABC Framework reverses this logic: governance must come first, ensuring that AI systems are deployed with clear frameworks of accountability, transparency, and oversight.

This is not theoretical. Regulatory developments such as the EU AI Act are reinforcing this shift, classifying applications in admissions, institutional assignment, and learning-outcome assessment as high-risk uses under Annex III, and positioning AI governance as an indicator of institutional quality and readiness. However, governance extends beyond compliance. It involves embedding ethical and operational considerations throughout the entire AI lifecycle.

In practice, accountable AI implementation includes:

• Establishing institution-wide norms while allowing contextual flexibility
• Defining risk-tier governance for high-stakes domains such as assessment and admissions
• Enabling participatory decision-making processes
• Building vetted access, procurement, and compliance infrastructures

In this sense, governance is not a constraint on innovation. It is the foundation that makes responsible innovation possible.

Bold: From pilots to institutional capability

If governance provides the foundation, boldness provides momentum. The ABC Framework emphasises that AI must be approached as a system-wide transformation, demanding coordinated, system-wide capability building. Bold implementation accounts for this scale and for the interdependencies across the system. In practice, this means treating institutional access, procurement, and compliance architecture as the foundation for capability building. A live inventory of AI systems, data flows, and vendors gives the institution real-time oversight. Boldness, therefore, requires structure and discipline.

Key elements of a bold approach include:

• Embedding AI into long-term institutional strategy, not treating it as a side initiative
• Making AI and data literacy a core competency for all stakeholders
• Establishing continuous cycles of experimentation, evaluation, and adaptation

Importantly, boldness is not about rapid disruption for its own sake. It is about building repeatable, scalable capabilities, allowing institutions to adapt over time.

Creative: Re-centring education on human judgment

The most profound implication of AI lies in its involvement in learning itself. As AI is integrated into decision-making and knowledge work, what remains distinctly human is judgment, problem-framing, critical scrutiny, and ethical responsibility.

The Creative pillar of the ABC Framework embraces this creativity by focusing on human judgment as the core of education. AI has revealed weaknesses in traditional educational proxies, such as output-based assessment. When producing the correct answer becomes easier, the value of education shifts towards the processes behind those answers: reasoning, interpretation, and decision-making.

This requires a rethinking of both teaching and assessment:

• Redesign assessment to reduce detection-dependence, increase process, oral defences, iterative work and transparent norms
• Develop student capabilities that balance disciplinary depth, technical literacy, critical thinking, and socio-emotional skills
• Support faculty in designing AI-inclusive learning situations and deciding when AI supports vs. short-circuits learning

In this AI-driven world, teaching cannot be simply the transmission of knowledge. It must be learning experiences that develop judgment and the ability to navigate complexity.

Applying the ABC Framework for Responsible AI in Higher Education

Balancing the three pillars

The effectiveness of the ABC Framework lies in the balance between its components. Accountability without Boldness becomes bureaucratic paralysis; Boldness without Accountability becomes uncontrolled diffusion. Creative redesign only works once both pillars are in place.

Institutions that align all three pillars are better positioned to move from fragmented initiatives to coherent transformation. This alignment enables them to respond to both technological change and the evolving expectations from employers, students, and society.

What institutional leaders should do next

Three places to start:

• Accountable: build a live inventory of AI deployments and data flows.
• Bold: require AI and data literacy across staff and students.
• Creative: redesign two high-stakes assessments per programme toward process and oral defence.

For the full set of priorities, see the AIHES Summit proceedings.

Designing the future of higher education in an AI-driven world

Higher education has historically played a central role in shaping economic and societal development. The current AI-driven transformation raises a familiar but urgent question: can institutions adapt at the pace and scale required?

The ABC Framework offers one possible approach. It is a collective coordination tool drawing upon the expertise and experimentation of educators around the world. The summit drew engaged adopters, and the framework should be read as a working hypothesis open to revision by institutions earlier in the AI journey. In the spirit of its message, this framework will be questioned, debated and challenged in the months and years to come. But by aligning governance, capability-building, and pedagogical redesign, it provides a shared language for institutions seeking to navigate this transition.

As accreditation and ranking bodies continue to redefine measures of quality and readiness, such a framework and language may help bring greater clarity and impact for our learners and our sector.

As a European business school, ESCP offers this framework as part of Europe’s response to responsible AI in higher education. AI transformation is institutional, human and strategic. It is a transformation that no single institution can navigate alone, which is why we launched the AI in Higher Education Summit and will reconvene with academic leaders again in 2027 to continue the conversation.

The ABC Framework is a starting point, not a finished solution, but the core premise is clear. The future of higher education depends on striking a balance between accountability, boldness and creativity.

To move from principles to practice, explore the full Executive Report from the AI in Higher Education Summit. It brings the ABC Framework to life through real institutional strategies, faculty experiments, and concrete use cases from across 67 universities worldwide.