ORM-2.2
A university or school training a model on it, under ORM 2.2
Educational institution, teaching or research use. Any machine learning training, fine-tuning, or distillation using the Work. This resolves to the Root tier with a Compute licence required on top.
Tier
Owed
No royalty, but a Compute Fee before the training run. The Root tier does not include training at any size.
Sections in play
1.7, 2.1, 2.2, 6.1, 6.2, 6.3, 6.5
Who and what
The situation being resolved.
Actor
A university or school
Educational institution, teaching or research use.
any
Situation
Training a model on it
Any machine learning training, fine-tuning, or distillation using the Work.
Rulings
4 questions resolved against the text.
Which tier applies?
Root. Educational institution, teaching or research use. Below $20,000,000 in trailing revenue, or within an exempt category, no Canopy royalty applies. Sections 4, 6, 8, and 10 still bind every Licensee.
Sections 1.7, 2.1, 2.2
Is anything owed?
No royalty, but a Compute Fee before the training run. The Root tier does not include training at any size.
Sections 6.1, 6.2
Does training need its own licence?
Yes, at every tier including Root. Section 6.1 removes AI Training Use from the base grant entirely. The fee is charged per training run consuming the weights or their outputs. Training without it is not a lesser breach of these terms, it sits outside the licence, which Section 6.5 treats as infringement.
Sections 6.1, 6.2, 6.5
What must the resulting model do?
Carry provenance disclosure on output that reproduces the Work substantially word for word, above the threshold set in Exhibit B. Section 6.3 targets the attribution harm alleged against Copilot rather than attempting to police memorisation.
Sections 6.3
Cautions
What this situation gets wrong most often.
Revenue is measured across entities under common control, per Section 1.4. A parent company above the threshold makes its subsidiary a Canopy Licensee regardless of that subsidiary's own turnover.
Corpus Inclusion under Section 1.11 reaches datasets compiled with actual knowledge they will be used for training. Passing the Work to a third party who trains on it does not move the obligation off you.
Section 8.3 forbids imposing a user ceiling, a regional exclusion, or an output-training restriction downstream where this licence imposes none. That prohibition is the point of the instrument.
Same actor
Other situations for this party.
Non-normative. Where this and the licence text disagree, the licence text applies.