AI

The AI coding tutor paradox grows as educators scramble to rethink how they test real skills

Featured from AI & Machine Learning Desk

An international survey of 763 computer science educators from 49 countries, conducted by an ACM task force, reveals that 68% of respondents have changed how

The AI coding tutor paradox grows as educators scramble to rethink how they test real skills
Gera Cejas

Key takeaways

  • 68 percent of surveyed computer science educators have adjusted their testing methods due to AI.
  • 64 percent of surveyed educators have shifted teaching away from writing code from scratch toward code comprehension, debugging, and problem-solving.
  • The top concern among educators is growing student dependency on technology (87 percent), followed by cheating and plagiarism (72 percent).
  • 48 percent of educators cite a lack of proven best practices as the biggest barrier to integrating AI.
  • An Anthropic study found that software developers who learned with AI assistance scored 17 percent lower on a follow-up knowledge test than those without AI access.

An international survey of 763 computer science educators from 49 countries, conducted by an ACM task force, reveals that 68% of respondents have changed how they test students and 64% have altered how they teach due to generative AI. Educators are shifting away from writing code from scratch toward code comprehension and debugging, while adopting proctored in-person exams, oral defenses, and paper-based tests to combat cheating. However, 48% of educators report a lack of proven best practices for integrating AI into their courses.

Watch the brief

In their words

“The key is to produce graduates who are competent in programming and who understand both the capabilities and limitations of AI.”
Steven Gordon, Professor at Ohio State University

By the numbers

68%
Educators who adjusted how they test students
763
Computer science educators surveyed globally
87%
Educators concerned about student dependency on technology
-17%
Knowledge test score drop for AI-assisted learners
-20%
Closed-book exam score drop after six months of AI use

How it unfolded

  1. May to October 2025 ACM task force surveyed 763 global educators

Turn stories like this into views

Ravenclip finds the AI news, makes the video, and posts it before attention moves on.

Start my channel

Source: The Decoder

Common questions

What happened with ACM?
68 percent of surveyed computer science educators have adjusted their testing methods due to AI.
Where can I read the original report?
Read the full report at the_decoder.

More in AI