TALAMANA · THE AI LITERACY MAP FOR ARCHITECTURE AND DESIGN · Ethics & Provenance · AGE 18—21 · FACTUAL · EVOLVING

What the licence lets you do

Only the text you agreed to says what you may do with an output.

When to use

Before any AI-made output goes into something you will publish, sell, or hand to a client. Not after.

The method

Four steps, in order. Find the terms for the exact model or service you used — the licence file for open weights, the terms of use for an API or app. Find the output clause: what it says about who owns outputs and whether commercial use is allowed, and any conditions — attribution, revenue thresholds, restricted uses. Find the input clause: what the service may do with what you uploaded. Record it next to the deliverable: service, date, the sentence that permits the use. Terms change; the date is part of the record.

Watch for this

Confusing you own the output with the output is copyrightable. A service can assign you whatever rights it has; it cannot create rights the law does not give. In the US, the Copyright Office holds that purely machine-generated material is not registrable. Indian law has a line for computer-generated works — Section 2(d)(vi) of the Copyright Act 1957: the author is "the person who causes the work to be created" — but whether a generative model's output is "computer-generated" in that sense is untested in Indian courts. Ownership and copyright are two different questions. And a second confusion: open weights are not open source. Published weights under a restrictive licence are still restricted; the Open Source Initiative's definition needs the freedom to use, study, modify and share, with enough of the code, the weights and the data information to do so.

Try it

For one model you use, find whether its terms allow commercial use of outputs and point to the exact sentence. Then compare two services — one open-weight, one paid — and show that your answer for each came from the licence text, not from the price or the download.

Prove it

For two models you use, one open-weight and one paid, state whether their terms allow commercial use of outputs and where the terms say so; and explain why the download and the price told you nothing the text did not.

How it works

Open-weight models ship with licences that range from fully permissive to use-restricted, and some add conditions above a certain scale of use; an open-weight licence can forbid commercial use. Paid services usually say in their terms that you own what you generate, subject to their policies; a free hosted service can grant the same. The variable is the text you agreed to. The licence covers your contract with the provider. Copyright covers your claim against everyone else. People mix the two up all the time. This card is EVOLVING because the terms change without notice and the copyright question is in court. Where a reference image or a dataset also carries a licence — the Creative Commons family is the most common — that licence travels with whatever you make from it.

What this idea builds on

What this idea opens up

Sources

Open this idea on the map · The complete map · Logika · RBDS AI Lab, India · revised every edition.

Age grows from 11 at the centre to 22 at the edge, and six sectors show the learning strands. Tab into the map and the arrow keys step from idea to idea, following the links where there is one. Enter opens the idea under the cursor, and E reads out its links and the reason recorded on each. Press slash for Search, question mark for the full key list, and Escape to leave. Open Ideas for the complete readable list, including what each idea builds on and what it opens up.

LOGIKA · RBDS AI LAB INDIA
ON-RAMP · AGE 11 · FIRST ENCOUNTERS, NOT GATES — IDEAS · — DEPENDENCIES
DONE
OPENS NEXT
SOLID — STANDS ON