docs: license summaries are not legal advice

The teacher-license notes read as legal conclusions. Add explicit
disclaimers to the README, the distillation guide, and the
permissive-teachers results: the descriptions are lay summaries as of
mid-2026, licenses change between versions, application depends on use,
and compliance is the user's responsibility to verify against the exact
weights they ship, with counsel where it matters.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

Author: mstrathman

README.md

@@ -241,8 +241,16 @@ TabPFN-prefixed model name) when you distribute the student, while the
 later TabPFN-2.5/2.6/3 weights are non-commercial. Your own labels or
 models carry no such limits. The
 extension's license gate (`accept_license`) makes a restrictively-licensed
-teacher an explicit opt-in rather than an accident, but checking that your
-use fits the teacher's license is on you.
+teacher an explicit opt-in rather than an accident.
+
+> [!IMPORTANT]
+> The license descriptions above are our lay summaries as of mid-2026,
+> not legal advice. Model licenses change between versions and over
+> time, and how they apply depends on your use. Before distilling any
+> third-party model or shipping a student derived from one, read the
+> license that came with the exact weights you are using, and consult
+> your own counsel where it matters. Compliance is your responsibility,
+> not something this tool or its documentation can determine for you.
 
 Three student kinds: `'tree'` is a single decision tree (a few kilobytes,
 interpretable); `'gbt'` is a gradient-boosted forest with second-order

benchmarks/results/permissive-teachers.md

@@ -1,9 +1,10 @@
 # Permissive tabular teachers: TabICL and Mitra
 
-Every teacher in this table may be distilled and the student
-shipped commercially: TabICL checkpoints are BSD-3-Clause, Mitra
-is Apache-2.0, TabPFN-2 is under the Prior Labs License
-(distillation permitted with attribution). Same datasets, splits,
+Every teacher in this table carried a permissive license when
+measured (mid-2026): TabICL checkpoints BSD-3-Clause, Mitra
+Apache-2.0, TabPFN-2 the Prior Labs License (distillation permitted
+with attribution). These are lay summaries, not legal advice;
+verify the license on the exact weights you use. Same datasets, splits,
 and caps as `tabarena-full.md`; xgboost/TabFM/TabPFN columns come
 from the other campaigns.
 

website/src/content/docs/guides/distillation.md

@@ -87,5 +87,13 @@ non-commercial license does both; TabPFN-2's Prior Labs License permits
 distillation, commercially too, with attribution when you distribute the
 student (the later TabPFN-2.5/2.6/3 weights are non-commercial); Chronos
 is Apache-2.0; your own labels and models carry no such limits. The `accept_license` option makes a
-restrictively-licensed teacher an explicit opt-in, but confirming that your
-use fits the teacher's terms is your responsibility.
+restrictively-licensed teacher an explicit opt-in.
+
+:::caution[Not legal advice]
+The license descriptions here are lay summaries as of mid-2026. Model
+licenses change between versions and over time, and how they apply
+depends on your use. Read the license that shipped with the exact
+weights you use, and consult your own counsel where it matters.
+Compliance is your responsibility; neither this extension nor its
+documentation can determine it for you.
+:::