Computer science educators are changing how they teach and test programming because AI coding tools can now solve many standard assignments. A survey by an ACM task force found that 69 percent of respondents believe AI has changed the skills needed for software development.
The survey covered more than 700 educators in 49 countries, with about 500 nearly complete responses analyzed. Sixty-four percent said they had already changed their teaching, often moving away from writing code from scratch and toward code comprehension, debugging, and problem-solving.
Assessment is changing even faster. Sixty-eight percent reported adjustments such as proctored in-person exams, oral exams, code-defense sessions, and project-based work. The goal is to measure whether students understand what code does and can reason about failures, not only whether they can produce a working snippet.
The transition is uneven. Nearly half of respondents said they lack proven examples for integrating AI into courses. That leaves schools balancing two needs at once: preparing students for AI-assisted workplaces while still ensuring they build durable programming judgment.