Embracing AI: What Are University Leaders Concerned About?

University leaders find themselves concerned about various aspects of integrating AI into education. To assist university leadership and AI developers in navigating the challenges facing higher education, DEC has conducted a comprehensive investigation into the top concerns shared by universities.

According to our research, these challenges and concerns can be broadly categorised into four main areas:

‍• Potential Negative Impact: This includes potential risks of incorporating AI into education, such as student data privacy issues, an increased risk of plagiarism, and the potential devaluation of degrees due to decreased exclusivity of high-quality education and over-reliance on AI.

• Implementation & Operations: This involves the complexity of implementing AI solutions, from choosing the appropriate AI solutions and integrating AI into the existing systems to providing training for faculty and ensuring students have equal access to AI.

• Organisational Structure: Universities often face difficulty with structural change and can be challenged by complex decision-making processes. However, we recommend universities consider how some structural adjustments may assist in more rapid responses to technological changes hastened by AI.

• External Environment: Beyond the institutions themselves, the direction and clarity of regulatory rules and guidelines governing the use of AI can significantly influence university AI strategies. However, with only a limited number of jurisdictions, such as the European Union and China having clear legislation on the use of AI in education, most nations, along with their higher education institutions, are still navigating uncertainty.

‍

Our research indicates that the potential negative impacts, along with the complexities of implementation and operation, are the primary concerns for most universities. We have identified five challenges that universities share as their  biggest concerns:

‍Ethical Concerns

AI solutions can collect extensive personal data from students, such as  academic records, behaviour patterns, and learning progress. Safeguarding student data and ensuring responsible data harvesting, storage, and disclosure are important challenges universities face.

Another notable ethical concern is around the explainability of Al, which comprises three aspects:

‍1. The trustworthiness of AI-driven insights, the university's ability to comprehend AI algorithms and explain the decisions made by AI.
2. Some output from AI is still unable to be explained or is simply wrong.
3. Ownership and copyright issues relating to the data that the AI tools were originally trained on are ongoing market concerns and subject to legal action with unknown outcomes and consequences.

‍Integration Complexity

Universities operate with many different existing systems, often from multiple vendors. Each system may have its own architecture, data formats, and integration requirements.  Universities often struggle with integrating systems and integrating new AI-based systems with legacy systems that pre-date AI could be fraught.

Plagiarism Risk

Plagiarism is  a persistent challenge and AI rapidly increases the volume, ease, and access to this old scourge. AI tools enable very high-quality academic misconduct, while AI detection tools lack sufficient accuracy and reliability. Rather than investing in AI detection tools that become obsolete almost as quickly as they are implemented, we recommend universities build a Resilient Educational Value Chain. (link to blog article)

‍Degree Devaluation

Degree devaluation is the concern of disruption to the higher education market, particularly among top-tier universities where the prestige of a degree relies on its exclusivity. As one top-tier university puts it, “we give a small number of students a badge and that badge has enormous value because we're highly selective”.  AI has the potential to disrupt this over the medium term by enabling the delivery of high-quality, personalised education, at scale and at a very low cost. Consequently, some universities are hesitant to fully embrace AI, as stated by the same university “we don't really have the motivation to adopt AI thoroughly”.

To address the challenge of degree devaluation, we believe universities should consider   business model innovation to unlock the growth of universities. DEC will soon share a framework to assist universities with their business model innovation process.

‍Tech Vendor Uncertainty

Currently, the EdTech market consists of  thousands of AI solutions and vendors, the most novel of which are from startups and early-stage companies. This poses a challenge for universities in selecting appropriate AI -based solutions, especially given the potential risk of errors from introducing new solutions and AI’s early-stage reliability challenges. 

Many universities have mentioned the urgent need for a curated selection of validated AI solutions to ease this selection process. There is no one-size-fits-all solution.

‍

—

The Digital Education Council has access to the world’s largest EdTech network, with over 10,000 solutions.

Members of the  Digital Education Council may refer to the Technology Scouting and Consultation team for bespoke advice on clearly defining AI-based technology needs and successful testing and implementation strategies. 

‍

‍