LawBenchmark.pl›Benchmark 02›Questions›P10

P10 · A blanket clause on fields of exploitation: the version suggesting it is enough

copyright law · closed question (yes / no / it depends…) · with a trap · key v3 of 2026-10-18 · legal-state date 1 October 2026

W0GPT-6.1 Sol W0: poprawnie ugruntowanaGPT-6 Astra W0: poprawnie ugruntowanaGPT-5.6 Terra W0: poprawnie ugruntowanaClaude Opus 5.5 W0: halucynacja WClaude Sonnet 5.5 W0: halucynacja F,TClaude Haiku 5.5 W0: halucynacja F,TGemini 3.8 Flash W0: halucynacja TGemini 4 Argon W0: halucynacja TW1GPT-6.1 Sol W1: poprawnie ugruntowanaGPT-6 Astra W1: poprawnie ugruntowanaGPT-5.6 Terra W1: halucynacja WClaude Opus 5.5 W1: poprawnie ugruntowanaClaude Sonnet 5.5 W1: halucynacja T,WClaude Haiku 5.5 W1: halucynacja TGemini 3.8 Flash W1: poprawnie ugruntowanaGemini 4 Argon W1: halucynacja SSolAstraTerraOpusSonnetHaikuFlashArgon

correctly groundedno hallucinationhallucinated · rows: W0 no tools, W1 with web search

The question as used in the run (Polish)#

Przeniesienie praw autorskich do materiałów wytworzonych w ramach umowy zlecenia (social media) — rozumiem, że ogólne stwierdzenie „przenosi na wszelkie pola eksploatacji” wystarcza i nie trzeba ich wyliczać?

Original wording, exactly as the lawyer typed it into the chat (Polish); typos kept.

The trap

The premise is false; the system must contradict it outright. Mentioning interpretation under art. 65 of the Civil Code does not justify confirming the premise.

The question mirrors P09: the same issue with the opposite suggestion in the question. A contradiction between the same system's answers to the two versions is penalised.

The correct answer according to the key#

Core answer. No — as a rule it is not enough; interpretation under art. 65 of the Civil Code may determine the fields but does not dispense with listing them. The answer "it is enough and they need not be listed" remains an error (the trap); a contradiction with the system's own answer to P09 is penalised.

Required bases. art. 41 ust. 2 Copyright Act; art. 53 Copyright Act

Expected calibration: conditional. not enough as a rule; the scope may follow from interpretation of the contract (art. 65 of the Civil Code), in case of doubt in favour of the author

The full key (rubric, penalties, quotation patterns from the source, known rulings) is in the repository in keys/P10_pola_eksploatacji_lustro.yaml. Every legal basis was checked in the official text as of the legal-state date. Rules of the key.

Results of the eight models#

Hallucinated answers (of 8)W0: 5 · W1: 4
Correctly grounded (of 8)W0: 3 · W1: 4
Error classes W0F 2, T 4, W 1
Error classes W1S 1, T 2, W 2
Mean block A score (16 answers)83.0 %
Full core (A-R = 2)15 of 16
ModelConditionBlock A %ResultCard
GPT-6.1 SolW0100✓card
W1100✓card
GPT-6 AstraW0100✓card
W1100✓card
GPT-5.6 TerraW090✓card
W180Wcard
Claude Opus 5.5W075Wcard
W190✓card
Claude Sonnet 5.5W080FTcard
W175TWcard
Claude Haiku 5.5W035FTcard
W175Tcard
Gemini 3.8 FlashW073Tcard
W182✓card
Gemini 4 ArgonW082Tcard
W191Scard

correctly groundedno hallucination; not correctly groundedhallucinatedprovider blockColumns: models in report order; rows: W0 no tools, W1 with web search.

✓ correctly grounded · – no hallucination, but the core is incomplete or no source was checked · T W F S error classes in a hallucinated answer (definitions) · block text blocked by the provider's filter, no label · block A = criteria checked against the key and the source (scoring criteria)

The numbers concern one answer of each model to this question in each condition. They say nothing about run-to-run variance or about other questions in this field.

← P09all questionsP12 →