68% преподавателей информатики изменили экзамены из-за ИИ — опрос ACM в 49 странах

The Association for Computing Machinery (ACM) published a survey in July 2026 of 763 computer science instructors from 49 countries: 68% of them have already changed their exam format because of generative AI, shifting toward oral exams, proctoring, and project-based work.

What the ACM survey found

The ACM survey recorded a mass overhaul of knowledge assessment: more than two-thirds of computer science instructors have restructured their exams to account for the widespread availability of AI assistants. This reflects a fundamental shift in emphasis — from the ability to write code to the ability to read, understand, and explain it.

• 763 computer science instructors from 49 countries
• 68% have already changed their exam format because of AI
• New formats: oral exams, proctoring, project-based work
• Almost half of respondents have no proven examples of integrating AI into their courses
• The focus of teaching is shifting from writing code to understanding it

Why instructors are changing their exams

Instructors are changing their exams because classic written coding assignments can now be solved by AI in seconds. Homework and standard tests no longer reflect a student's real level: tools like ChatGPT, Claude, and Copilot write working functions from a short prompt, and copying a solution has become easier than writing it yourself.

In response, 68% of those surveyed switched to formats that are harder to delegate to a model. An oral exam shows whether a student understands their own solution; proctoring restricts access to AI during the work; a project defense checks the ability to bring a system to a working result and explain every decision made. In essence, universities are returning to live dialogue with the student in places where automated code checking used to be enough.

The paradox of the AI tutor

The paradox is that the very same AI that devalues old exams turns out to be the strongest personal programming tutor in history. It explains mistakes, breaks down algorithms, and offers step-by-step hints — but at the same time it allows someone to get through a course without ever learning to write code on their own.

That is precisely why instructors are hastily rethinking how to test real skills. And here the main problem surfaces: almost half of the ACM survey's participants admit they have no proven examples of how to correctly build AI into the learning process. Right now, essentially, every university and every course is experimenting alone, without a shared methodological foundation.

Teaching is shifting from writing code to understanding it. —

conclusion of the ACM survey among 763 instructors from 49 countries

What this means

Higher education in computer science is being restructured faster than guidelines can appear: two-thirds of instructors have already changed their exams, but there is still no generally accepted recipe for how to teach programming in the age of AI. In the coming years, universities will be searching for formats where AI helps people learn rather than replacing self-learning, and the ACM survey is the first major measurement of how quickly this is happening.

Frequently asked questions

How many instructors took part in the ACM survey?

The ACM survey involved 763 computer science instructors from 49 countries. It is one of the largest international snapshots of opinion on how AI is affecting programming education.

How are universities changing their exams because of AI?

According to ACM, 68% of instructors switched to oral exams, proctoring, and project-based work. These formats are harder to get through with the help of AI and better test understanding of code rather than the ability to generate it.

What is preventing AI from being built into courses?

Almost half of instructors said they have no proven examples of integrating AI into teaching. Ready-made methods and standards for using models without harming skills have not yet taken shape.