Between the holiday break and the start of regular business hours on January 3, 2024, Cambridge University Press & Assessment and Cognizant discreetly announced that they had extended a multi-year collaboration, this time with an increased emphasis on artificial intelligence. As is always the case, the announcement was made through a press release that was easy to scroll past due to its measured corporate language. However, the scope of what was described merits a closer examination.
Cambridge is not your typical publisher. It is a part of the University of Cambridge, one of the world’s oldest and most prestigious academic institutions, and it has a truly vast impact on classrooms, serving millions of students in dozens of countries through standardized tests, English language programs, and international education systems. When a company of that size enters into a five-year contract with a large technology services company to promote digital transformation and integrate generative AI into its operations, the consequences typically spread in ways that no press release can adequately convey.
In addition to continuing to develop and support applications, Cognizant, a Nasdaq-listed business with a strong foundation in enterprise technology services, will assist Cambridge in adopting what the announcement refers to as a “product-centric approach.” It’s worthwhile to consider the term “product-centric.” In the past, academic publishing has structured itself around content, including books, curricula, syllabi, and exam frameworks developed over many years. A product-centric model iterates based on platform data and user behavior, thinking more like a software company. It is not merely a technical change, but a true philosophical one.
The agreement’s specific objectives include enhancing student platform speed and interactivity, preserving the integrity of exam results, and utilizing generative AI to produce what both businesses refer to as “productivity improvements” and improved learning outcomes. Regarding exam integrity, there is real cause for optimism—AI-assisted monitoring and verification tools have become significantly more advanced, and technology can conceivably assist in resolving the problem of maintaining fair, accurate assessments on a global scale. It is more difficult to assess the learning outcomes component. Everyone in EdTech seems to be promising personalized, adaptive experiences these days, but there is still a stubbornly large gap between what students actually see on a screen and what they are promised.
It’s difficult to ignore the fact that Cambridge has purposefully framed its AI strategy around human oversight, at least in public. The organization has declared that it wants AI to assist teachers rather than compete with them. This commitment may seem simple, but as commercial pressures increase, it will take true institutional discipline to uphold. Additionally, Cambridge has released guidelines for educational institutions regarding generative AI in coursework and preventing plagiarism, indicating that it is at least carefully considering the long-term consequences of the technology it incorporates into its own products.
The most honest thing to say as we watch this partnership develop is that there is genuine uncertainty about the outcome. The organizations in question are reputable. The intentions seem thoughtful. However, results from multi-year AI contracts in education often differ from the press release at year five. In the end, it will be up to the students taking exams based on Cambridge frameworks—in Singapore, Kenya, the UK, and Chile—to determine whether this was successful.
