TALAMANA · THE AI LITERACY MAP FOR ARCHITECTURE AND DESIGN · Ethics & Provenance · AGE 16—19 · FACTUAL · EVOLVING
New forms of dishonesty — not all of them plagiarism
A false claim of authorship is the offence, whichever name it gets.
The idea
The old rule was simple: do not hand in someone else's work as yours. That is plagiarism — presenting another person's words, ideas or work as your own, without credit. AI adds cases that are dishonest without being plagiarism. A generated image captioned as a hand drawing is misrepresentation. A model's whole scheme handed in where the brief required your own hand is unauthorised assistance. A made-up source is fabrication. Each is a false claim about how the work was made, so one test covers them: does the reader believe you did something you did not do?
Why it matters
Every one of these is a false claim about who made the work, or how. The tool has changed; the false claim has not. The word matters too: an integrity panel charges a named offence, and "it was not plagiarism" is no defence against misrepresentation.
See it in the studio
A portfolio page says hand-rendered perspective. It was generated, then traced. The tracing took skill. The caption is still false, because the caption is about who made the image, not how good it looks. The right name for it is misrepresentation, not plagiarism — and it carries the same penalty.
Watch for this
"It's not plagiarism because no person was copied." Fair — it may not be. It may be misrepresentation, fabrication, or a breach of the assessment rules, and those are offences too. A false claim of authorship is the offence, whichever name it gets.
Try it
List five ways AI could touch one studio project, from a spell-check to a full generated scheme. For each, write what caption would be honest, and what the dishonest caption would be called — plagiarism, misrepresentation, fabrication, unauthorised assistance. Notice where the honest caption becomes embarrassing — that is the line.
Prove it
Name three specific acts with AI that would count as academic dishonesty, give each its right name, and say why each is a false claim about authorship. Then explain why a generated image shown as a hand drawing is misrepresentation, however good it looks.
How it works
Institutions are rewriting their integrity codes around these cases, and the wording differs from school to school and year to year — which is why the card is EVOLVING. Most codes now separate the categories: plagiarism, fabrication, undisclosed or unauthorised assistance, prohibited delegation, false process claims, and copyright infringement, which is a matter of law rather than of the code. They overlap, and one act can be two of them. The idea that lasts is older than any code: when you submit work, you are claiming something about how it was made. The Declaring AI card on this map is the rule to find first; this card is what the rule is protecting.
What this idea builds on
What this idea opens up
Sources
- ETH.A
- social-impact-ethics
- UNESCO, AI competency framework for students
Open this idea on the map · The complete map · Logika · RBDS AI Lab, India · revised every edition.