AI use in this book is governed by the purpose of each task, not by one blanket rule. Every assessed activity uses one of three labels.
AI-free¶
Complete the task without generative AI assistance. You may use the book, supplied examples, instructor-approved references, and ordinary editor features.
This mode protects practice that you need to perform personally: reading short programs, tracing values, recalling core structures, interpreting elementary errors, and establishing an individual baseline.
Unless a task says otherwise, do not paste the prompt into an AI tool, request a solution, use generated completion, or submit generated explanations in an AI-free task.
AI-supported¶
AI may support learning within stated limits—for example by offering a hint, posing a question, explaining an error message, or commenting on code you already attempted.
The task will state what is allowed. Keep the first attempt, relevant prompt or interaction summary, changes made, and a brief explanation of what you accepted or rejected.
AI-integrated¶
Using AI is part of the task. You may be asked to design a prompt, compare proposals, test generated code, identify unsupported claims, improve a solution, or document risks and limitations.
A generated artifact is never sufficient evidence by itself. You remain responsible for the specification, verification, attribution or disclosure, and final result.
What you always owe the reader or assessor¶
Whatever the mode, submit honest evidence of your work. Depending on the task, that evidence may include:
the requirement or specification;
your prediction or first attempt;
code and data versions;
tests and observed results;
revisions and their reasons;
an assistance disclosure; and
a short explanation or live defense.
Do not enter confidential, personal, licensed, or restricted material into an AI service unless you have explicit authorization and understand the service’s data terms. Do not fabricate a process record. Do not treat an AI detector as proof of authorship; assessment should rely on meaningful evidence and explanation.
The agreement¶
As a learner, you agree to use assistance in the declared mode, preserve enough evidence to explain your process, and take responsibility for what you submit.
As an author or instructor, we agree to state the mode and rationale clearly, assess the published outcomes, avoid surprise restrictions, and design work in which understanding is more valuable than imitation.