rights audit: add Sejm SIS terms citation pass (2026-10-05)

The SIS commercial-offer clause dates from early 2016, is scoped to
www.sejm.gov.pl, and is narrower than a commercial-use ban. Update the
parliamentary, sejm_api*, committee and interpellations rows and cards.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

artifacts/rights_audit_2026-10/REVIEW.html

@@ -438,7 +438,7 @@ tr.grp td{background:var(--bg);font-weight:600;font-family:var(--f-display)}
   <article class="src part" id="parliamentary">
     <header><h3>parliamentary</h3><span class="pill p-part">Partial</span><span class="rel">v0.2.5</span><span class="reg">release 0.2.0 · registry <code>CC-BY-4.0</code>, card <code>public-domain</code> · SpeakLeash <code>PPC_corpus</code></span></header>
     <div class="grid">
-      <div><h4>Problem</h4><p>The chain is Sejm, then IPI PAN PPC, then SpeakLeash, then us. The only rights statement is IPI PAN's "public domain", which is a redistributor's claim. Committee sittings dominate the text and have speaker labels stripped, so guests can't be filtered out. The registry and the card disagree on the licence.</p></div>
+      <div><h4>Problem</h4><p>The chain is Sejm, then IPI PAN PPC, then SpeakLeash, then us. IPI PAN's "public domain" is a redistributor's claim with no stated basis. The Sejm website's own SIS terms (since early 2016) ask for a source note and bar SIS material from the commercial offer of an agency, TV station or portal. Committee sittings dominate the text and have speaker labels stripped, so guests can't be filtered out. The registry and the card disagree on the licence.</p></div>
       <div><h4>Evidence</h4><ul>
         <li><span class="tag">V</span> 324,622 docs, by heuristic classification:
           <ul>
@@ -448,6 +448,8 @@ tr.grp td{background:var(--bg);font-weight:600;font-family:var(--f-display)}
           </ul>
         </li>
         <li><span class="tag">V</span> Samples quote constituents by name. Some items predate 1991 and some have mojibake. There are no per-doc dates.</li>
+        <li><span class="tag">V</span> SIS terms: SIS material "nie mogą być elementem komercyjnej oferty żadnej agencji, stacji telewizyjnej, portalu ani innego podmiotu". Wayback shows only a permissive sentence on 2015-11-21; the clause first appears on 2016-02-04. SIS is the www.sejm.gov.pl system (Regulamin Sejmu art. 201a).</li>
+        <li><span class="tag">I</span> The clause is narrower than a commercial-use ban and says nothing about datasets. Open Data Act art. 15 ust. 1 lists the conditions a public body may set, and a commercial bar is not among them. PPC crawl dates are unknown, so some text predates the clause. Details: <a href="reports/citations/sejm_sis.md">sejm_sis.md</a>.</li>
         <li>Classifier: <code>official/ppc_classify.py</code>.</li>
       </ul></div>
       <div class="full"><h4>Fix</h4><ol>
@@ -501,6 +503,7 @@ tr.grp td{background:var(--bg);font-weight:600;font-family:var(--f-display)}
           </ul>
         </li>
         <li><span class="tag">V</span> Non-official examples include Konfederacja Lewiatan, Grupa Azoty S.A., Greenpeace, ClientEarth, HFPC, Związek Miast Polskich and university professors. There are 3,097 "Głos z sali" (voice from the floor) turns.</li>
+        <li><span class="tag">I</span> Regulamin Sejmu art. 166 ust. 4 puts the full record of committee sittings in SIS, so the SIS source-note and commercial-offer terms may apply. Scope and enforceability: <a href="reports/citations/sejm_sis.md">sejm_sis.md</a>.</li>
         <li>Role shares: <code>official/committee_roles.csv</code>. Classifier: <code>committee_speakers.py</code>. Known false positives: NBP, Wody Polskie, IPN (all state bodies).</li>
       </ul></div>
       <div class="full"><h4>Fix</h4><ol>
@@ -683,7 +686,7 @@ tr.grp td{background:var(--bg);font-weight:600;font-family:var(--f-display)}
   <article class="src minor" id="sejm_api">
     <header><h3>sejm_api</h3><span class="pill p-minor">Minor</span><span class="rel">v0.2.5</span><span class="reg">release 0.2.5 · <code>LicenseRef-Polish-Official-Documents</code> · <code>src/fetch_sejm_api.py</code></span></header>
     <div class="grid">
-      <div><h4>Problem</h4><p>The data comes from the HF snapshot <code>PiotrSty/sejm-speeches-corpus@124557799608</code>, not the live API. <code>source_url</code> points only at a day-level <code>/videos/</code> endpoint. The API publishes no licence or terms. Content is MPs' and ministers' floor speeches: 38,812 rows from 1,149 authors.</p></div>
+      <div><h4>Problem</h4><p>The data comes from the HF snapshot <code>PiotrSty/sejm-speeches-corpus@124557799608</code>, not the live API. <code>source_url</code> points only at a day-level <code>/videos/</code> endpoint. The API publishes no licence or terms. The SIS terms cover www.sejm.gov.pl, and no source puts API output in SIS. Content is MPs' and ministers' floor speeches: 38,812 rows from 1,149 authors.</p></div>
       <div><h4>Fix</h4><p>Change provenance to "Sejm API data via contributor snapshot". Spot-check 20 rows against the live API. Use per-speech URLs if the API offers them.</p></div>
     </div>
   </article>
@@ -691,7 +694,7 @@ tr.grp td{background:var(--bg);font-weight:600;font-family:var(--f-display)}
   <article class="src minor" id="sejm_api_2011_2022">
     <header><h3>sejm_api_2011_2022</h3><span class="pill p-minor">Minor</span><span class="reg">unreleased · <code>LicenseRef-Polish-Official-Documents</code> · <code>src/fetch_sejm_api_2011_2022.py</code></span></header>
     <div class="grid">
-      <div><h4>Problem</h4><p>Same snapshot chain as <code>sejm_api</code>: 118,902 rows from 2,095 authors. NERGAL hasn't run. It overlaps <code>parliamentary</code> (2011–2019) and the registered <code>parlamint_pl</code>.</p></div>
+      <div><h4>Problem</h4><p>Same snapshot chain as <code>sejm_api</code>: 118,902 rows from 2,095 authors. Terms 7–8 (2011–2015) predate the 2016 SIS wording. NERGAL hasn't run. It overlaps <code>parliamentary</code> (2011–2019) and the registered <code>parlamint_pl</code>.</p></div>
       <div><h4>Fix</h4><p>Apply the <code>sejm_api</code> fix. Dedup against <code>parliamentary</code> and <code>parlamint_pl</code> before release. Run NERGAL.</p></div>
     </div>
   </article>
@@ -703,8 +706,8 @@ tr.grp td{background:var(--bg);font-weight:600;font-family:var(--f-display)}
         <li>2.2% quote 400+ chars.</li>
         <li>2.9% quote constituents.</li>
         <li>About 4% match a "Pan/Pani Name" pattern.</li>
-      </ul><p>PDF-only replies are skipped.</p></div>
-      <div><h4>Fix</h4><p>Add <code>license-evidence.json</code> with a caveat about third-party quotations. Run NERGAL before release, as the card requires.</p></div>
+      </ul><p>PDF-only replies are skipped. Regulamin art. 192 puts interpellations and replies in SIS, so the SIS terms may apply; replies are written by ministries, not the Kancelaria. The card's <code>added</code> date (2026-09-05) disagrees with <code>sources.py</code> (2026-07-19).</p></div>
+      <div><h4>Fix</h4><p>Add <code>license-evidence.json</code> with a caveat about third-party quotations and the SIS terms. Fix the <code>added</code> date. Run NERGAL before release, as the card requires.</p></div>
     </div>
   </article>
 

artifacts/rights_audit_2026-10/reports/citations.md

@@ -1,6 +1,6 @@
 # Rights audit by release (2026-10-05)
 
-Class = audit severity (REVIEW.html). Cited = primary-source citation pass of 2026-10-05: **V** read first-hand, **—** not re-cited (audit evidence only; HPLT-class and Partial rows are in scope). Full citation tables: [`nkjp_novels.md`](citations/nkjp_novels.md), [`fandom.md`](citations/fandom.md), [`parliamentary.md`](citations/parliamentary.md), [`openstax_cv.md`](citations/openstax_cv.md), [`samorzad_nonsa.md`](citations/samorzad_nonsa.md), [`wikimedia.md`](citations/wikimedia.md). `raw/...` paths in them name local evidence captures (Wayback pages, API responses, PDFs) that are not committed.
+Class = audit severity (REVIEW.html). Cited = primary-source citation pass of 2026-10-05: **V** read first-hand, **—** not re-cited (audit evidence only; HPLT-class and Partial rows are in scope). Full citation tables: [`nkjp_novels.md`](citations/nkjp_novels.md), [`fandom.md`](citations/fandom.md), [`parliamentary.md`](citations/parliamentary.md), [`openstax_cv.md`](citations/openstax_cv.md), [`samorzad_nonsa.md`](citations/samorzad_nonsa.md), [`wikimedia.md`](citations/wikimedia.md), [`sejm_sis.md`](citations/sejm_sis.md). `raw/...` paths in them name local evidence captures (Wayback pages, API responses, PDFs) that are not committed.
 
 ## v0.2.5 (tag 7ae4ac7, 20 sources)
 
@@ -12,7 +12,7 @@ Class = audit severity (REVIEW.html). Cited = primary-source citation pass of 20
 | wiktionary_examples | 8,639 | HPLT-class | CC-BY-SA-3.0 | [Szablon:NKJP/opis](http://web.archive.org/web/20260619101211/https://pl.wiktionary.org/wiki/Szablon:NKJP/opis), [Szablon:źródło](http://web.archive.org/web/20260908194648/https://pl.wiktionary.org/wiki/Szablon:%C5%BAr%C3%B3d%C5%82o); recompute on `wk_pages.json` | V | Stats don't reproduce: 73 entries, 109 lines; 45.0% of lines / 60.0% of cleaned chars cited (56% of entries); NKJP 20.2% of raw chars. `{{źródło}}` is a generic formatter; `{{NKJP}}` is dropped unrecorded; "skip `<ref`" fix over-broad. No pl.wiktionary policy on NKJP examples found. Fetcher L41 labels all rows 3.0 (wrong for post-cutoff text) | Rebuild keeping uncited examples only |
 | 1000_novels | 1,000 | HPLT-class | CC-BY-4.0 | CLARIN-PL [11321/312](http://hdl.handle.net/11321/312) `dc.rights` CC BY 4.0, no third-party carve-out; WL XML `dc:rights`; [mahajana.net/node/414](http://mahajana.net/node/414) (BY-ND 3.0); Friedman d. 2006 [Q11778742](https://www.wikidata.org/wiki/Q11778742) | V | "Covers only the compilation" is our reading, not CLARIN's. 3 Hōjun texts are BY-SA 3.0 Unported at WL but BY-ND at origin, so relabelling to BY-SA is unsupported | Drop 17 + 143, or relabel 17 (amended: relabel only the 14 PL ones; withhold the 3 Hōjun) |
 | nkjp1m | 18,381 | HPLT-class | CC-BY | NKJP book (PWN 2012) [§5.3 "Dostępność", p. 57](http://nkjp.pl/settings/papers/NKJP_ksiazka.pdf); [clip.ipipan](https://clip.ipipan.waw.pl/NationalCorpusOfPolish), [nkjp.pl p14](http://nkjp.pl/index.php?page=14&lang=1), [HF ipipan/nkjp1m](https://huggingface.co/datasets/ipipan/nkjp1m) | V | "Not a grant" → "no publisher grant documented". Licences differ by version (GPL v3 = 1.0/1.1; CC-BY = 1.2), not a conflict. Card line "covers the corpus text" has no source | Remove, or get IPI PAN written confirmation |
-| parliamentary | 324,622 | Partial | CC-BY-4.0 (card: PD) | [IPI PAN PPC](https://clip.ipipan.waw.pl/PPC): "parliamentary data is public domain", annotations CC-BY, no basis stated; Copyright Act [art. 4](https://eli.gov.pl/api/acts/DU/2025/24/text/O/D20250024.pdf) (Dz.U. 2025 poz. 24); [Sejm SIS terms](http://web.archive.org/web/20260603115729/https://sejm.gov.pl/Sejm10.nsf/page.xsp/copyright) (archive; live page CAPTCHA) | V | Conflict is inside the registry (`sources.py:197` PD vs `:198` CC-BY-4.0); card repeats both (`:9`, `:14`). "PPC under CC-BY-4.0" unsupported (CC-BY is annotations only). Not the only statement: **Sejm SIS terms require attribution and bar use "w komercyjnej ofercie"** (reach to API/PDF text unverified). `created` 1991–2019 contradicted (doc from 1967). Shares re-counted: committee 10.7% docs / 58.6% chars | Drop committee docs (59% of chars) or the source |
+| parliamentary | 324,622 | Partial | CC-BY-4.0 (card: PD) | [IPI PAN PPC](https://clip.ipipan.waw.pl/PPC): "parliamentary data is public domain", annotations CC-BY, no basis stated; Copyright Act [art. 4](https://eli.gov.pl/api/acts/DU/2025/24/text/O/D20250024.pdf) (Dz.U. 2025 poz. 24); [Sejm SIS terms](http://web.archive.org/web/20260603115729/https://sejm.gov.pl/Sejm10.nsf/page.xsp/copyright) (archive; live page CAPTCHA) | V | Conflict is inside the registry (`sources.py:197` PD vs `:198` CC-BY-4.0); card repeats both (`:9`, `:14`). "PPC under CC-BY-4.0" unsupported (CC-BY is annotations only). Not the only statement: **Sejm SIS terms** (in force since early 2016; absent on 2015-11-21) ask for the source note "System Informacyjny Sejmu" on material from www.sejm.gov.pl and bar SIS material from being "elementem komercyjnej oferty" of an agency, TV station or portal; not a general commercial-use ban, silent on datasets. SIS = www.sejm.gov.pl only (Regulamin art. 201a); no terms on api.sejm.gov.pl; enforceability doubtful under Open Data Act art. 15 ust. 1 (closed list of conditions, no commercial bar). PPC crawl dates unknown, so some text predates the bar ([Sejm SIS](citations/sejm_sis.md)). `created` 1991–2019 contradicted (doc from 1967). Shares re-counted: committee 10.7% docs / 58.6% chars | Drop committee docs (59% of chars) or the source |
 | samorzad_gov_pl | 72,771 | Partial | CC-BY-SA-4.0 | Footer [govpl_template.js](https://web.archive.org/web/20260929062416/https://samorzad.gov.pl/scripts/govpl_template.js?locale=pl_PL) (script-injected, unchanged since 2026-08-13); [gov.pl/prawa-autorskie](https://web.archive.org/web/20261002174202/https://www.gov.pl/web/gov/prawa-autorskie) (carve-out, names Minister Cyfryzacji); onboarding guide PR-CYBER-06 (Dec 2020, JST); IMGW-PIB notice on gmina pages + [danepubliczne.imgw.pl/regulations](https://web.archive.org/web/20250910211311/https://danepubliczne.imgw.pl/regulations) | V | **No tenant licence grant found**: the two KPRM agreements the guide names are unpublished. Footer names no licensor ("jest udostępniane"); operator named inconsistently (MC vs KPRM). IMGW is not a ban but "wyłącznie w formie dosłownej" + attribution, incompatible with BY-SA (9 docs). "112 pupil/participant/poem docs" not reproducible (127 titles with "wiersz", 69 Sokołów). 2–3% third-party share stays I | Drop third-party subsets; confirm with KPRM/MC |
 | wikipedia | 1,171,897 | Minor | CC-BY-SA-3.0 | [WMF ToU oldid 554823](https://foundation.wikimedia.org/w/index.php?title=Policy:Terms_of_Use&oldid=554823) (BY-SA 4.0 in effect 2023-06-07; [talk oldid 261297](https://foundation.wikimedia.org/w/index.php?title=Policy_talk:Terms_of_Use&oldid=261297)) | V | Cutoff 2023-06-01 is 6 days early, not conservative: use 2023-06-07 (06-08 if time of day unknown). Technical footer-switch date unchecked | Rebuild from official dump; label 3.0/4.0 |
 | wikinews | 24,386 | Minor | CC-BY-2.5 | [Warunki użycia oldid 342796](https://pl.wikinews.org/w/index.php?title=Wikinews:Warunki_u%C5%BCycia&oldid=342796) ("CC BY 4.0 od 16 grudnia 2024"); [T381421](https://phabricator.wikimedia.org/T381421); [Gerrit 1101867](https://gerrit.wikimedia.org/r/c/operations/mediawiki-config/+/1101867) (synced 08:32 UTC) | V | Switch is 2024-12-16, not the vote window (2024-07-22 to 08-20). Card labels all rows 2.5; whether any rows postdate the switch is unchecked | Label 2.5/4.0 by edit date |
@@ -23,7 +23,7 @@ Class = audit severity (REVIEW.html). Cited = primary-source citation pass of 20
 | eurlex | 243,060 | Minor | CC-BY-4.0 | — | — | | Relabel as EU reuse policy |
 | govpl | 88,190 | Minor | CC-BY-SA-4.0 | — | — | | Restore third-party carve-out; drop PAP docs; add URLs |
 | global_voices | 2,040 | Minor | CC-BY-3.0 | — | — | | Drop 64 partner-republished docs |
-| sejm_api | 38,812 | Minor | Polish official documents | — | — | | State the intermediary snapshot |
+| sejm_api | 38,812 | Minor | Polish official documents | [Sejm SIS](citations/sejm_sis.md); Regulamin Sejmu (M.P. 2026 poz. 573) art. 201a | V | No terms in API docs, headers or sampled payloads; no source puts API output in SIS. Text came via the PiotrSty HF release; `source_url` is a `/videos/{date}` listing. NOTICE credits "Kancelaria Sejmu RP, api.sejm.gov.pl" without the SIS wording | State the intermediary snapshot |
 | dziennik_ustaw | 35,442 | Clean | CC0-1.0 | — | — | | Fix card date range |
 | eltec_pol | 100 | Clean | CC-BY-4.0 | — | — | | Add authors to card; dedup vs wolne_lektury |
 
@@ -34,12 +34,12 @@ Class = audit severity (REVIEW.html). Cited = primary-source citation pass of 20
 | openstax_pl | HPLT-class | CC-BY-4.0 | 8 live [/pages/przedmowa](https://openstax.org/books/psychologia-polska/pages/przedmowa) blocks (2026-10-05); [Psychologia archive 2026-01-19](https://web.archive.org/web/20260119202244/https://openstax.org/books/psychologia-polska/pages/przedmowa); [CMS API](https://openstax.org/apps/cms/api/v2/pages/?type=books.Book&fields=title,license_name,license_version,license_url,book_state&limit=200) | V | Timeline is staged, not "older than our snapshot": AI-training ban archived since 2024-04-16 (beside "CC BY"); site footer flipped to NC-SA between 2026-04-11 and 06-06; long LLM wording, book-level NC and CMS NC-SA first seen live 2026-10-05 (archived CMS records say CC BY 4.0). Live block contradicts itself (label "(CC BY)", CC BY deed link). Irrevocability of earlier CC BY stays I | Pull: AI-training ban predates our fetch; NC now stated |
 | common_voice_pl | HPLT-class | CC0-1.0 | CV [europarl file @8f2f7c0](https://github.com/common-voice/common-voice/blob/8f2f7c077a1f67bc539a8a7951c4c64a2c349cb2/server/data/pl/selected-europarl-v7-pl.txt) ([commit](https://github.com/common-voice/common-voice/commit/8f2f7c077a1f67bc539a8a7951c4c64a2c349cb2)); [sentence-collector.txt @99a4910](https://github.com/common-voice/common-voice/blob/99a49105a77aa9e120aec6fe1f764e3672f0f3f8/server/data/pl/sentence-collector.txt); [EP legal notice](https://www.europarl.europa.eu/legal-notice/en/) (unchanged 2020–2026); EP Bureau Decision [C/2025/341](https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=OJ:C_202500341) art. 4; [statmt.org/europarl](http://www.statmt.org/europarl/) | V | Counts reproduce only with NFKC + casefold + drop non-word chars (stated recipe gives different numbers; fix the spec). "Entire item" overstated: partial reuse with item URL allowed. EP open data is CC BY 4.0 since 2025 (not CC0; coverage of 1996–2011 text unestablished). "WL excerpts, translators never checked" unsupported; replace with "Sentence Collector lines are volunteer-attested PD; sources not recorded" | Withdraw, or drop the 24,709 Europarl sentences |
 | fandom_pl | HPLT-class (per wiki) | CC-BY-SA-3.0 | [Fandom ToU](https://web.archive.org/web/20260905223208/https://www.fandom.com/terms-of-use) rev. 2025-12-19, "User conduct" (scraping, automated access, robots.txt, **AI-training ban**); [fandom.com/licensing](https://www.fandom.com/licensing); [Bulbapedia:Copyrights](https://bulbapedia.bulbagarden.net/wiki/Bulbapedia:Copyrights) (BY-NC-SA 2.5); CDPR [fan content](https://www.cdprojektred.com/en/fan-content) + [user agreement](https://regulations.cdprojektred.com/en/user_agreement) (no TDM) | V | Not an intermediary: `PiotrSty/fandom-pl` is the preparer's own repo, live API scraped 2026-10-02 after the ToU revision. "Bulk reuse" is not ToU wording. All 12 wikis report CC-BY-SA with no version | Drop 6 wikis, sample 4; ToU AI clause likely drops the whole source |
-| sejm_committee_transcripts | Partial | Polish official documents | Copyright Act [art. 4](https://eli.gov.pl/api/acts/DU/2025/24/text/O/D20250024.pdf) (Dz.U. 2025 poz. 24); Open Data Act [art. 6 ust. 2, ust. 4 pkt 4](https://eli.gov.pl/api/acts/DU/2023/1524/text/O/D20231524.pdf) (Dz.U. 2023 poz. 1524); SN V CSK 337/08 (2009, "public-task body" test); [Sejm SIS terms](http://web.archive.org/web/20260603115729/https://sejm.gov.pl/Sejm10.nsf/page.xsp/copyright) (archive; live page CAPTCHA) | V | Reading "partly": art. 4 pkt 2 is about documents, not speakers; ODA art. 6(4)(4) bites only where a third party holds copyright; art. 6 ust. 2 (personal data of private persons) is the stronger hook for guests. No ruling on guests at committee sittings found. Re-counted: 653,411 turns, non-official 5.01% chars, 3,097 "Głos z sali" | Keep only official speakers |
+| sejm_committee_transcripts | Partial | Polish official documents | Copyright Act [art. 4](https://eli.gov.pl/api/acts/DU/2025/24/text/O/D20250024.pdf) (Dz.U. 2025 poz. 24); Open Data Act [art. 6 ust. 2, ust. 4 pkt 4](https://eli.gov.pl/api/acts/DU/2023/1524/text/O/D20231524.pdf) (Dz.U. 2023 poz. 1524); SN V CSK 337/08 (2009, "public-task body" test); [Sejm SIS terms](http://web.archive.org/web/20260603115729/https://sejm.gov.pl/Sejm10.nsf/page.xsp/copyright) (archive; live page CAPTCHA) | V | Reading "partly": art. 4 pkt 2 is about documents, not speakers; ODA art. 6(4)(4) bites only where a third party holds copyright; art. 6 ust. 2 (personal data of private persons) is the stronger hook for guests. No ruling on guests at committee sittings found. Re-counted: 653,411 turns, non-official 5.01% chars, 3,097 "Głos z sali". Regulamin art. 166 ust. 4 puts full committee records in SIS, so the SIS terms may reach this text (inference; [Sejm SIS](citations/sejm_sis.md)). Card cites ODA art. 15, which limits conditions rather than grants reuse | Keep only official speakers |
 | nonsa | Partial | CC-BY-SA-3.0 | [siteinfo rightsinfo](https://nonsa.pl/api.php?action=query&meta=siteinfo&siprop=rightsinfo&format=json) (CC BY-SA 3.0); [Prawa autorskie oldid 1850937](https://nonsa.pl/index.php?title=Nonsensopedia:Prawa_autorskie&oldid=1850937); [dump 2025-10-09](https://archive.org/details/wiki-nonsa.pl-20251009); [HF PiotrSty/nonsa-pl @09f62dc](https://huggingface.co/datasets/PiotrSty/nonsa-pl/blob/09f62dc3826159dbc8c0df7a95243a18ec52abda/README.md) | V | ns100 Cytaty confirmed (287 docs, 2.61% tokens) but it is film/game/music *quotes*; "lyrics (Metallica, Acid Drinkers)" unsupported. 12 `<poem>`/0.5% stays I (2 checked: credited band verse). 100/100 dump ids gone live; sample of 5 deleted by vote, none for copyright. `ns` already in attribution sidecar, missing from parquet. Card sentence is local only (`nonsa.md:32`); HF has 12,391 vs local 12,304 docs (gap unexplained). Site-level licence only; per-page tags unchecked | Drop ns100 "Cytaty" and the `<poem>` docs |
 | plwiki_talk | Minor | CC-BY-SA-4.0 | [WMF ToU oldid 554823](https://foundation.wikimedia.org/w/index.php?title=Policy:Terms_of_Use&oldid=554823) (BY-SA 4.0 in effect 2023-06-07; [talk oldid 261297](https://foundation.wikimedia.org/w/index.php?title=Policy_talk:Terms_of_Use&oldid=261297)) | V | Cutoff 2023-06-01 is 6 days early, not conservative: use 2023-06-07 (06-08 if time of day unknown). Technical footer-switch date unchecked | Label per record by revision date |
 | plwiki_user_talk | Minor | CC-BY-SA-4.0 | [WMF ToU oldid 554823](https://foundation.wikimedia.org/w/index.php?title=Policy:Terms_of_Use&oldid=554823) (BY-SA 4.0 in effect 2023-06-07; [talk oldid 261297](https://foundation.wikimedia.org/w/index.php?title=Policy_talk:Terms_of_Use&oldid=261297)) | V | Cutoff 2023-06-01 is 6 days early, not conservative: use 2023-06-07 (06-08 if time of day unknown). Technical footer-switch date unchecked | Label per record by revision date |
-| sejm_api_2011_2022 | Minor | Polish official documents | — | — | | As sejm_api; dedup; NERGAL |
-| sejm_interpellations | Minor | Polish official documents | — | — | | Add evidence file; NERGAL |
+| sejm_api_2011_2022 | Minor | Polish official documents | [Sejm SIS](citations/sejm_sis.md) | V | As sejm_api. Terms 7–8 (2011–2015) predate the 2016 SIS wording; Sejm6/Sejm7 terms not found | As sejm_api; dedup; NERGAL |
+| sejm_interpellations | Minor | Polish official documents | [Sejm SIS](citations/sejm_sis.md); Regulamin art. 192, 193, 195 | V | Regulamin puts interpellations, answers and written questions in SIS. Answers are ministry-authored, so the Kancelaria is a weaker rights-holder (inference). No NOTICE. `added` conflicts: `sources.py` 2026-07-19 vs card 2026-09-05 | Add evidence file; NERGAL |
 | uzp_orzeczenia_pl | Minor | Polish official documents | — | — | | State review status; dedup with SAOS |
 | ted_pl | Minor | TED notice reuse | — | — | | Cite both bases; full name/email scan |
 | kultura_bez_barier_pl | Minor | CC-BY-SA-3.0 | — | — | | Note version gap; PII in one PDF |

artifacts/rights_audit_2026-10/reports/citations/parliamentary.md

@@ -123,7 +123,7 @@ C1. Audit: "registry CC-BY-4.0, card public-domain". Wrong in where the conflict
 
 C2. Registry / card: "PPC redistributed by IPI PAN under CC-BY-4.0" is unsupported. IPI PAN states: "The parliamentary data is public domain. The corpus annotations are available on CC-BY (attribution) licence." No CC version, and CC-BY applies to annotations only (P1-P3). Replacement: "IPI PAN states the parliamentary data is public domain and offers its annotations under CC-BY (version unstated); no legal basis is given."
 
-C3. Audit: "The only rights statement is IPI PAN's 'public domain'". Incomplete: (a) SpeakLeash's manifest repeats "Public Domain" (derived, K1); (b) the Sejm's own SIS page sets terms: free use with attribution to "System Informacyjny Sejmu" and a bar on inclusion in any entity's commercial offer (T2-T4). Whether those terms reach the API/biuletyn PDFs and the PPC text is not established (T5). "Public domain" is a redistributor's claim with no stated basis: supported (P4).
+C3. Audit: "The only rights statement is IPI PAN's 'public domain'". Incomplete: (a) SpeakLeash's manifest repeats "Public Domain" (derived, K1); (b) the Sejm's own SIS page sets terms: free use with attribution to "System Informacyjny Sejmu" and a bar on inclusion in any entity's commercial offer (T2-T4). Whether those terms reach the API/biuletyn PDFs and the PPC text is not established (T5). "Public domain" is a redistributor's claim with no stated basis: supported (P4). Narrowed by [`sejm_sis.md`](sejm_sis.md): the bar dates from early 2016 (absent on 2015-11-21), covers a "commercial offer" of an agency, TV station or portal rather than commercial use in general, SIS is the www.sejm.gov.pl system only, and enforceability under Open Data Act art. 15 ust. 1 is doubtful.
 
 C4. Audit: "api.sejm.gov.pl/sejm.html states no licence or terms of use (V)". True for that page only (T1). Replacement: "api.sejm.gov.pl/sejm.html states no licence; the SIS page on sejm.gov.pl states reuse terms (T2-T4)."
 

artifacts/rights_audit_2026-10/reports/citations/sejm_sis.md

@@ -0,0 +1,134 @@
+# Sejm "System Informacyjny Sejmu" (SIS) terms: scope, history, channel reach, enforceability
+
+Retrieved 2026-10-05. V = read first-hand this session (page, PDF, header, file). I = inference or judgement; no primary source states it. Quotes are verbatim Polish, at most 25 words; "..." marks an elision. Raw evidence is in `scratchpad/citations/raw/` (prefix `sis_`; some rows cite the earlier-pass `parl_` files). Paths in the "source" column are relative to that directory unless they start with `polish-dynaword/`. Wayback copies were reached only through existing snapshots and CDX listings (no Save Page Now). One HTTP 429 round on the availability API: one 100 s back-off, then recorded as "not checked".
+
+Access notes: www.sejm.gov.pl and orka.sejm.gov.pl return an F5 "Human Verification" page to bots (not bypassed); the terms page was read from Wayback. ISAP is blocked and was not used; statutes came from the Sejm ELI API and Dziennik Ustaw PDFs.
+
+---
+
+## 1. Claim table
+
+### 1a. Wording of the SIS terms (Sejm10 page)
+
+| id | claim | V/I | source | quote |
+|---|---|---|---|---|
+| W1 | The Sejm10 page "System Informacyjny Sejmu" (`/Sejm10.nsf/page.xsp/copyright`) grants free use of SIS materials "with the restrictions below". Latest readable copy: Wayback 2026-06-03. | V | `parl_wayback_sejm10_copyright_20260603.html` (one-line HTML; search the text). Archive: http://web.archive.org/web/20260603115729/https://sejm.gov.pl/Sejm10.nsf/page.xsp/copyright | "Zezwala się na bezpłatne wykorzystywanie materiałów znajdujących się w Systemie Informacyjnym Sejmu, z ograniczeniami przedstawionymi poniżej" |
+| W2 | Attribution clause is scoped to material "pobieranych ze strony www.sejm.gov.pl" (taken from the website). | V (wording); the scope reading is I | same file | "przy wszystkich materiałach (video, audio czy tekstowych) pobieranych ze strony www.sejm.gov.pl powinna znajdować się informacja o źródle pochodzenia" |
+| W3 | Required source wording. | V | same file | "Źródłem tym jest System Informacyjny Sejmu" |
+| W4 | Commercial bar. Not limited to iTV Sejm material ("w szczególności"), not channel-limited in its text, says nothing of datasets or model training. | V (wording); breadth is I | same file | "materiały zamieszczone w SIS, w szczególności ... iTV Sejm, nie mogą być elementem komercyjnej oferty żadnej agencji" |
+| W5 | The page itself says SIS is "available via the Internet at www.sejm.gov.pl" and is for gathering/processing/publishing data on the Sejm, its organs and the Kancelaria. | V | same file | "System Informacyjny Sejmu jest dostępny poprzez sieć Internet pod adresem http://www.sejm.gov.pl/" |
+| W6 | The same page quotes the Access to Public Information Act: public information is subject to reuse "on the rules and in the manner set by the Act". | V | same file | "podlega udostępnieniu i ponownemu wykorzystywaniu na zasadach i w trybie określonych w ustawie" |
+| W7 | The page contains no reference to the API, ELI, orka, PDF or text/data mining. | V (negative, text search of the 2026-06-03 copy) | same file | n/a |
+
+### 1b. History (Wayback, existing snapshots)
+
+| id | claim | V/I | source | quote |
+|---|---|---|---|---|
+| H1 | Sejm8 page, 2015-11-17: permissive wording only. Free copying with source indication; no commercial bar; no "iTV Sejm". | V | `sis_wayback_sejm8_copyright_20151117.html`. http://web.archive.org/web/20151117194033/http://sejm.gov.pl/Sejm8.nsf/page.xsp/copyright | "Zezwala się na wykorzystywanie (w tym kopiowanie) materiałów znajdujących się w Systemie Informacyjnym Sejmu w sieci Internet, z zaznaczeniem źródła ich pochodzenia" |
+| H2 | Sejm8 page, 2015-11-21 20:09 UTC: same permissive wording; no "komercyjnej", no "iTV". Last capture without the bar. | V | `sis_wayback_sejm8_20151121200928_copyright.html`. http://web.archive.org/web/20151121200928/http://sejm.gov.pl/sejm8.nsf/page.xsp/copyright | same sentence as H1 |
+| H3 | Sejm8 page, 2016-02-04 18:06 UTC: first capture with the three-clause text and the commercial bar. The page was also redesigned (title "System Informacyjny Sejmu", definition sentence added). | V | `sis_wayback_sejm8_20160204180642_copyright.html`. http://web.archive.org/web/20160204180642/http://www.sejm.gov.pl:80/Sejm8.nsf/page.xsp/copyright | "nie mogą być elementem komercyjnej oferty żadnej agencji, stacji telewizyjnej, portalu ani innego podmiotu" |
+| H4 | Bar present in every later capture read: Sejm8 20160222065114, 20160311235644, 20160413013006, 20170311093126; Sejm9 20191112004316 and 20211005175213; Sejm10 20260603115729. | V | `sis_wayback_sejm8_20160222065114_copyright.html`, `..._20160311235644_...`, `sis_wayback_sejm8_20160413_copyright.html`, `sis_wayback_sejm8_20170311_copyright.html`, `sis_wayback_sejm9_20191112_copyright.html`, `parl_wayback_sejm9_copyright_20211005.html`, `parl_wayback_sejm10_copyright_20260603.html` | n/a (string "komercyjnej oferty" present in each; grep count 1) |
+| H5 | The Sejm8 CDX lists no capture between 20151121200928 and 20160204180642. | V | `sis_cdx_sejm8.txt` (21 captures, first 20151117194033) | n/a |
+| H6 | So the commercial bar was introduced between 2015-11-21 and 2016-02-04. The exact date and the act that set the page text were not found. | I (bounds are V; "introduced in the window" is inference from captures) | H2, H3, H5 | n/a |
+| H7 | No act found that sets the page text. Zarządzenie nr 2 Szefa KS of 26.01.2016 (inside the window) amended Zarządzenie 3/2006, but the consolidated text shows it changed only the name of the unit handling requests (Biuro Komunikacji Społecznej) and § 6; it adds no reuse terms. | V (the text); no causal link claimed | `sis_zsk_2006_3_ujed4.txt` lines 1-125 (footnotes 3-11) | n/a |
+| H8 | Sejm7 and Sejm6 copyright pages: CDX returned nothing. The Sejm10 CDX could not be obtained (Archive "Temporarily Offline" page). Wayback copies of the Sejm10 page for 2025-01-17 and 2026-09-07 are bot-challenge pages with no terms text. | V | `sis_cdx_sejm7.txt`, `sis_cdx_sejm6.txt` (empty); `sis_cdx_copyright_pages.txt` (head); `parl_wayback_sejm10_copyright_20250117.html`, `..._20260907.html` | n/a |
+
+### 1c. Is "SIS" officially defined, and what is in it?
+
+| id | claim | V/I | source | quote |
+|---|---|---|---|---|
+| D1 | Regulamin Sejmu (current consolidated text M.P. 2026 poz. 573) art. 201a creates SIS as an electronic system for gathering, processing and publishing data on the Sejm, its organs and the Kancelaria. No technical or URL definition. | V | `sis_regulamin_sejmu_MP2026_573.txt:3495` | "Tworzy się System Informacyjny Sejmu, prowadzony w formie elektronicznej, przeznaczony do gromadzenia, przetwarzania i udostępniania danych" |
+| D2 | Art. 201a is absent from the 2002 consolidated text (M.P. 2002 poz. 398) and present in the 2009 text (M.P. 2009 poz. 47); SIS was created in the Regulamin between those two. | V (PDF text search for "201a" / "System Informacyjny": 2002 no, 2009 yes) | `sis_regulamin_MP_2002_398.pdf`, `sis_regulamin_MP_2009_47.pdf/.txt` | n/a |
+| D3 | Art. 202a ust. 1: the Kancelaria makes public information available by publishing it in SIS. | V | `sis_regulamin_sejmu_MP2026_573.txt:3539-3540` | "Informację publiczną Kancelaria Sejmu udostępnia poprzez ogłaszanie dokumentów i innych informacji w Systemie Informacyjnym Sejmu" |
+| D4 | Zarządzenie nr 3 Szefa Kancelarii Sejmu of 6.03.2006 (consolidated text) § 1 identifies SIS with the Sejm website. It is the only implementing act found; it contains no reuse, attribution or commercial terms. | V | `sis_zsk_2006_3_ujed4.txt:12-22` | "udostępniane są w Systemie Informacyjnym Sejmu ... na stronie internetowej Sejmu pod adresem http://www.sejm.gov.pl" |
+| D5 | Neither the Regulamin nor Zarządzenie 3/2006 mentions api.sejm.gov.pl, orka/orka2 or ELI. | V (negative; grep for "api.sejm", "orka", "eli.gov", "/eli" in both texts) | `sis_regulamin_sejmu_MP2026_573.txt`, `sis_zsk_2006_3_ujed4.txt` | n/a |
+| D6 | Committee full transcripts (art. 166 ust. 4) and subcommittee transcripts (art. 166a ust. 3) are to be placed in SIS. | V | `sis_regulamin_sejmu_MP2026_573.txt:2905, 2911` | "W Systemie Informacyjnym Sejmu, o którym mowa w art. 201a, umieszcza się wskazany w ust. 2 pełny zapis przebiegu obrad" |
+| D7 | Interpellation texts (art. 192 ust. 7), answers (art. 193 ust. 2) and, by cross-reference, written MP questions "zapytania poselskie" (art. 195 ust. 2) are published in SIS. | V | `sis_regulamin_sejmu_MP2026_573.txt:3297-3298, 3300-3303, 3350` | "teksty interpelacji, którym Prezydium Sejmu nadało bieg, są udostępniane w Systemie Informacyjnym Sejmu" |
+| D8 | Plenary stenograms: art. 176 makes the stenographic report the only official record of the sitting, but only the debate amendments are expressly placed in SIS (art. 176 ust. 2a). I found no express rule putting the full plenary stenogram in SIS; that it is there is a matter of practice (the Sejm site's stenogram pages, cited by the PPC paper, see C8). | V (articles); the practice conclusion is I | `sis_regulamin_sejmu_MP2026_573.txt:3074, 3081` | "które stanowią jedyne urzędowe stwierdzenie przebiegu obrad" |
+| D9 | The Regulamin is an internal resolution of the Sejm (uchwała), not a statute; it binds organs and the Kancelaria, not third-party users. | I | n/a | n/a |
+
+### 1d. Channels
+
+| id | claim | V/I | source | quote |
+|---|---|---|---|---|
+| C1 | The Sejm API docs (`/`, `/sejm.html`, `/eli.html`) contain no licence, terms of use or copyright statement. The only licence string is the Asciidoctor stylesheet header. The only contact line points to the Kancelaria. | V | `sis_live_api_root.html:470, 491`; `sis_live_api_sejm.html.html:11`; `sis_live_api_eli.html.html:11`; also `parl_api_sejm_gov_pl_sejm_html.html` | "Contact: The Chancellery of the Sejm" |
+| C2 | API responses for a committee PDF and a transcript HTML (2026-10-05) carry no licence, link, copyright or rights headers; `/robots.txt` is 404. | V | `sis_live_api_headers.txt` | `HTTP/1.1 404 Not Found` (robots.txt) |
+| C3 | Sample payloads carry no SIS, copyright, licence or commercial-offer text: committee PDF (ASW, term 10, sitting 1), plenary transcript HTML (term 10, proc. 1, 2023-11-13), interpellation body HTML (term 10, no. 1). Header on the PDF: "KANCELARIA SEJMU / Biuro Komisji Sejmowych". | V (3 samples; not all documents) | `sis_api_committee_ASW_1.pdf`, `sis_api_transcript_t10_p1_d1_1.html`, `sis_api_interp_t10_1_body.html`; earlier pass `parliamentary.md` T9 | n/a |
+| C4 | Sejm10 site footer links both "API" (https://api.sejm.gov.pl) and "System Informacyjny Sejmu" (the terms page). The bot-challenge page of www.sejm.gov.pl / orka says "For API requests use api.sejm.gov.pl or eli.gov.pl". Layout facts, not statements that the API is or is not part of SIS. | V (the text); significance is I | `sis_sejm10_informacja_publ.html` (footer, line 53); `parl_www_sejm_gov_pl_.html:128`; `parl_orka_sejm_gov_pl_Home_nsf_page_xsp_copyright.out:128` | "For API requests use api.sejm.gov.pl or eli.gov.pl" |
+| C5 | Whether API output is "in SIS" is stated nowhere I could read. Indirect pointers only: D6/D7 (document classes the Regulamin puts in SIS are served by the API), C4 (footer link). | I | D5, C1-C4 | n/a |
+| C6 | orka/orka2: terms not read. orka.sejm.gov.pl answers bots with the human-verification page; not bypassed. No repo fetcher uses orka (all use api.sejm.gov.pl). | V (block, fetchers) | `parl_orka_sejm_gov_pl_Home_nsf_page_xsp_copyright.out`; `polish-dynaword/src/fetch_sejm_*.py` | n/a |
+| C7 | dane.gov.pl: API search found no Kancelaria Sejmu institution (query "Sejm": 0 institutions; "Kancelaria": Kancelaria Senatu, KPRM) and no dataset for "api.sejm.gov.pl" (0 hits). Listing absence is not proof; no "warunki ponownego wykorzystywania" listing for Sejm data was found. | V (results); absence is I | `sis_danegov_inst_sejm.json`, `sis_danegov_inst_kancelaria_sejmu.json`, `sis_danegov_search_api_sejm_gov_pl.json`, `sis_danegov_search_Kancelaria_Sejmu.json`, others | n/a |
+| C8 | PPC's own stated source for plenary stenograms is the Sejm website: the LREC 2012 paper cites http://www.sejm.gov.pl/Sejm7.nsf/stenogramy.xsp and the orka2 search interface for the 7th term. The 2022 ParlaCLARIN paper names no channel and neither paper names any terms. | V | `parl_lrec2012_653.txt:10-26, 40-52`; `parl_parlaclarin2022_6.txt` (negative) | "Since 1993 their current versions are being made available online in the form of publicly available transcripts" |
+| C9 | Repo cards, NOTICEs, evidence JSON and fetchers for the five sources never use the words "System Informacyjny Sejmu" or "komercyjnej oferty". Sources named instead: "Kancelaria Sejmu RP, https://api.sejm.gov.pl/" (sejm_api, sejm_api_2011_2022, sejm_committee_transcripts NOTICEs; interpellations card "Source (upstream): https://api.sejm.gov.pl/"). `parliamentary` and `sejm_interpellations` have no NOTICE.md. | V | `polish-dynaword/data/<source>/` (grep, rc=1), `NOTICE.md` of the three; `polish-dynaword/src/sources.py:192-295` | `Source: Sejm committee full transcripts ... Kancelaria Sejmu RP, https://api.sejm.gov.pl/` |
+
+### 1e. Legal basis (all conclusions I unless marked V)
+
+| id | claim | V/I | source | quote |
+|---|---|---|---|---|
+| L1 | Open Data Act (tj. Dz.U. 2023 poz. 1524) art. 3 pkt 1: obliged bodies include "jednostki sektora finansów publicznych" (PFA); art. 3 pkt 2: other state organisational units without legal personality. | V | `sis_oda_DU2023_1524.txt:306-310` | "jednostki sektora finansów publicznych w rozumieniu przepisów ustawy z dnia 27 sierpnia 2009 r. o finansach publicznych" |
+| L2 | PFA (Dz.U. 2025 poz. 1483) art. 9 pkt 1 lists "organy władzy publicznej" and pkt 3 "jednostki budżetowe". | V | `sis_pfa_DU2025_1483.txt:444-449` | "organy władzy publicznej, w tym organy administracji rządowej" |
+| L3 | The Kancelaria Sejmu is an obliged body under art. 3. No primary source states this; none of art. 4 ust. 1 pkt 1-7 exclusions (public broadcaster/PAP, culture institutions, universities, schools, art. 3 pkt 5 entities, juvenile/prison units) names the Sejm or its Kancelaria. | I (art. 4 list: V) | `sis_oda_DU2023_1524.txt:361-392`; grep "sejm" in the ODA text hits only the promulgation block | n/a |
+| L4 | Art. 5 pkt 1: everyone may reuse information made available "w innym systemie teleinformatycznym podmiotu zobowiązanego" (so a Sejm system counts, with or without the BIP). Art. 2 pkt 12: reuse means use "w jakimkolwiek celu". | V | `sis_oda_DU2023_1524.txt:397-399, 298-299` | "wykorzystywanie przez użytkowników informacji sektora publicznego w jakimkolwiek celu" |
+| L5 | Art. 14 ust. 1: public sector information is made available for reuse unconditionally except as the Act provides. | V | `sis_oda_DU2023_1524.txt:546-547` | "udostępnia się lub przekazuje w celu ponownego wykorzystywania bezwarunkowo, z wyjątkiem przypadków określonych w ustawie" |
+| L6 | Art. 14 ust. 2: the body sets conditions for information that has features of a work, related-right subject matter or a database "do których przysługują mu prawa". This is the only route for conditions beyond art. 15, and it requires the Kancelaria to hold the rights. | V (text); reading is I | `sis_oda_DU2023_1524.txt:548-553` | "do których przysługują mu prawa, w szczególności zobowiązując użytkownika do poinformowania o nazwisku, imieniu lub pseudonimie twórcy" |
+| L7 | Art. 15 ust. 1 lists the conditions an obliged body may set: (1) naming source/time, (2) naming processing, (3) liability scope, (4) personal-data information. A bar on use in a commercial offer is not among them. | V (text); "not among them" is a V negative | `sis_oda_DU2023_1524.txt:554-565` | "dotyczące obowiązku poinformowania o źródle, czasie wytworzenia i pozyskania informacji sektora publicznego od podmiotu zobowiązanego" |
+| L8 | Art. 15 ust. 2 allows conditions that restrict "w działalności komercyjnej" only for museums, libraries and archives. The Kancelaria is none of these (the Biblioteka Sejmowa is a separate unit and is carved out in Zarządzenie 3/2006 § 6). | V (text); applicability is I | `sis_oda_DU2023_1524.txt:566-574`; `sis_zsk_2006_3_ujed4.txt` § 6 | "Muzea państwowe, muzea samorządowe, biblioteki publiczne, biblioteki naukowe, biblioteki pedagogiczne i archiwa mogą określić inne niż wymienione w ust. 1 warunki" |
+| L9 | Art. 16: conditions must be objective, proportionate, non-discriminatory and must not unjustifiably limit reuse. Art. 8 ust. 1: same terms in comparable situations. Art. 9 ust. 1: the body does not restrict use by other users (exclusive agreements only under art. 9 ust. 2, public-task necessity). Art. 17: reuse free of charge. | V | `sis_oda_DU2023_1524.txt:461-473, 579-585` | "Określone przez podmiot zobowiązany warunki ponownego wykorzystywania są obiektywne, proporcjonalne, niedyskryminacyjne" |
+| L10 | Art. 12 ust. 2-3: conditions set for information made available otherwise than via BIP/portal are an offer; reuse of information made available under art. 5 pkt 1 is acceptance of that offer. Art. 11 ust. 1 requires conditions to be published in BIP category "Ponowne wykorzystywanie". So a website notice can bind as an ODA offer, but only for conditions the Act allows (L5-L9). | V (text); effect is I | `sis_oda_DU2023_1524.txt:497-499, 524-531` | "stanowią ofertę zawierającą te warunki" |
+| L11 | The SIS commercial bar therefore does not fit the closed list in art. 15 ust. 1 and runs against arts. 14 ust. 1, 16 and 17. It is enforceable, if at all, only through rights the Kancelaria holds (art. 14 ust. 2): copyright (but see L13) or the database right (L12). Unverified by any authority; no commentary or case law read. | I | L5-L10 | n/a |
+| L12 | Database Act (DU 2024/1769; earlier text DU 2021/386): producer = who bears investment risk (art. 2 ust. 1 pkt 4); art. 6 ust. 1 exclusive right to extract/re-use all or a substantial part; art. 7 ust. 1 and 4: a lawful user cannot be barred from insubstantial parts for any purpose and contrary contract terms are void; art. 8 ust. 1 pkt 2: illustration/teaching/research use of a substantial part with source, if non-commercial; art. 8a: reproduction for text and data mining allowed unless the rightholder reserved otherwise (machine-readable reservation for publicly accessible online databases); art. 10: 15 years from creation. Whether the Kancelaria is a "producer" of a protected database (substantial investment) is not stated anywhere I read. The SIS commercial sentence is human-readable HTML text; no machine-readable reservation was found on api.sejm.gov.pl (C1, C2). Art. 8a covers reproduction for TDM, stored only as long as needed (ust. 3), not redistribution of a corpus. | V (statutory text); application is I | `sis_dbact_DU2024_1769.txt:45-62, 112-130, 147-175, 194`; `sis_dbact_DU2021_386.txt` | "Postanowienia umowne sprzeczne z ust. 1 i 2 są nieważne" |
+| L13 | Copyright: official documents and materials are outside copyright (Copyright Act art. 4 pkt 2, Dz.U. 2025 poz. 24; earlier pass `parliamentary.md` S1, S11). Committee and plenary records are by statute "urzędowe stwierdzenie przebiegu obrad" (art. 166 ust. 1; art. 176 ust. 1). If no copyright subsists, a notice cannot create an exclusive right in the text; the SIS bar would then rest on contract/ODA offer only (L10, L11). Whether guests' statements keep their own copyright is open (earlier pass S10, S14). | V (statutes); conclusions I | `sis_regulamin_sejmu_MP2026_573.txt:2896-2898, 3074`; `parliamentary.md` S1, S10, S11 | "stanowi urzędowe stwierdzenie przebiegu obrad" |
+| L14 | The SIS page is a unilateral website notice by the Kancelaria: not a statute, a regulation or a licence naming an ODA art. 15 ust. 3 standard open licence. Its binding force on someone who never visited www.sejm.gov.pl (API or intermediary user) is unproven. | I | W1, D4, C1 | n/a |
+| L15 | The attribution clause (source note, "Źródłem tym jest System Informacyjny Sejmu") fits ODA art. 15 ust. 1 pkt 1 and is the lawful kind of condition if it applies at all. The NOTICEs already name "Kancelaria Sejmu RP" and api.sejm.gov.pl (C9) but not the SIS wording. | I | L7, C9 | n/a |
+
+---
+
+## 2. Verdict per source
+
+"SIS clause applies?" covers both clauses. Short form: attribution clause = scoped to material "pobieranych ze strony www.sejm.gov.pl" (W2); commercial bar = "materiały zamieszczone w SIS" (W4).
+
+| source | channel | fetch / acquisition date | SIS clause applies? | basis |
+|---|---|---|---|---|
+| `sejm_committee_transcripts` | api.sejm.gov.pl `GET /sejm/term{n}/committees/{code}/sittings/{num}/pdf`, terms 9-10 (V: `polish-dynaword/src/fetch_sejm_committee_transcripts.py`; card L10-12). Derived via PiotrSty/sejm-committee-transcripts. | Sittings 2019-11-14 to 2026-09-03; evidence file "verified 2026-09-18"; `sources.py` added 2026-09-18. So fetched between about 2026-09-03 and 2026-09-18 (I). | Unclear; the strongest of the five. | Document class is named in Regulamin art. 166 ust. 4 as placed in SIS (D6), so the page text plausibly covers the same documents (I). But the PDF carries no terms (C3), the API docs carry none (C1), and nothing says API output is "in SIS" (C5). Clause was in force on the Sejm10 page at the last readable capture 2026-06-03 (W1); state on the fetch date unverified (H8). Enforceability doubtful (L7, L11). |
+| `sejm_interpellations` | api.sejm.gov.pl HTML `/body` endpoints for `interpellations` and `writtenQuestions`, term 10 (card L46-50; `sources.py:249-270`). | Records 2023-11-15 to 2026-09-03. Card says "Added 2026-09-05"; `sources.py` says `added` 2026-07-19 (the two disagree; V). Fetched on or after 2026-09-03 (I). | Unclear. | Question texts and answers are placed in SIS by art. 192 ust. 7, 193 ust. 2, 195 ust. 2 (D7). No terms in the sample body (C3). Answers are authored by ministries and other bodies, not by the Kancelaria Sejmu, so the Kancelaria is a weaker rights-holder for them (I; L6, L13). No NOTICE.md and no source-attribution text beyond the card line (C9). |
+| `sejm_api` | `PiotrSty/sejm-speeches-corpus@v1.0.1` (HF), which names its source as "Kancelaria Sejmu RP, API Sejmu" (V: `sis_hf_piotrsty_sejm_speeches_README_v1.0.1.md:90`). Per-row `source_url` is `https://api.sejm.gov.pl/sejm/term{n}/videos/{date}` (V: attribution.jsonl row 1). | Records 2023-01-11 to 2026-07-03; release date and crawl date not stated in the README; `sources.py` added 2026-08-22. So acquired after 2026-07-03 (I). | Unclear; weaker than committee/interpellations. | Intermediary dataset; repo never touched the Sejm site. `source_url` is a date-level `/videos/` listing, which does not identify the document class. If the text comes from plenary stenograms, placement in SIS is by practice only (D8), not by an express rule; if from iTV Sejm material, the named "iTV Sejm" wording in W4 would be closer to the source (I; method not stated by the README). |
+| `sejm_api_2011_2022` | Same release (`@v1.0.1`, commit 1245577 in `sources.py`; 124557799608 in attribution). `source_url` pattern same (`.../term7/videos/2011-11-08`). | Records 2011-11-08 to 2022-12-14; `sources.py` added 2026-08-31. | Unclear; same as `sejm_api`. | As `sejm_api`. Terms 7-8 content predates the clause on the Sejm pages if it was published before 2016 (H2-H3), but the repo copy was acquired in 2026 through an intermediary, so publication date does not decide anything (I). |
+| `parliamentary` | PPC (IPI PAN) via SpeakLeash `PPC_corpus` (V: `parl_speakleash_PPC_corpus.manifest.json`: licence "Public Domain", creation_date 2023-01-25, version 2.0). PPC page: "as of 22 May 2025". PPC's own source for plenary stenograms: Sejm site and orka2 (C8). | Data 1991-01-01 to 2019-12-31; SpeakLeash dataset created 2023-01-25; `sources.py` added 2026-07-02. IPI PAN crawl dates not stated. | Unclear; weakest direct link. | Repo got it from SpeakLeash/IPI PAN, which declare the parliamentary data "public domain" and apply CC-BY to annotations only (earlier pass P1-P3); no terms attach to that channel. The content derives from the Sejm site (C8), so the SIS terms could in theory travel with it, but: (a) the bar did not exist on the Sejm8 page before 2016-02 (H2, H3), and the share of PPC text crawled after that date is unknown; (b) composition per the earlier classification: 256,854 of 324,622 docs are interpellation questions/replies, 34,718 committee, 1,500 plenary, 31,550 other (`parl_ppc_classify_out.txt`; heuristic classifier). Repo wording "PPC redistributed by IPI PAN under CC-BY-4.0" is contradicted by IPI PAN (earlier pass P3). |
+
+Overall: no source can be said to be subject to the SIS clauses ("yes"), and none can be cleared as "no". All five are "unclear". Reasons: (1) V: no terms attach to the API channel, to its payloads, or to the intermediaries; (2) V: SIS is officially defined as the www.sejm.gov.pl system (D1, D4) and the Regulamin lists committee transcripts and interpellations among SIS content (D6, D7), so textual reach to the same documents is plausible (I); (3) I: the commercial bar is out of line with ODA arts. 14-17 and with the Database Act arts. 7 and 8, so enforceability is doubtful even where it reaches.
+
+---
+
+## 3. Corrections to the audit statement
+
+Audit statement: "Sejm SIS terms require attribution and bar use in a commercial offer (reach to API/PDF unverified)."
+
+1. "Sejm SIS terms": supported for the "System Informacyjny Sejmu" page on www.sejm.gov.pl (V, W1-W4; Sejm8 from 2016, Sejm9, Sejm10). The three-clause wording is not old: the 2015-11-17 and 2015-11-21 pages carried a permissive sentence and no commercial bar (V, H1-H2). The bar appears between 2015-11-21 and 2016-02-04 (H3, H6).
+2. "require attribution": the clause is scoped to material "pobieranych ze strony www.sejm.gov.pl" and prescribes the source "System Informacyjny Sejmu" (W2, W3). Scoped to the website, so its reach to API downloads is the same open question as the bar.
+3. "bar use in a commercial offer": the text is narrower than a general commercial-use ban. It bars SIS material from being "elementem komercyjnej oferty" of an agency, TV station, portal or other entity, with iTV Sejm broadcasts as the named example (W4). It says nothing of datasets, open releases or training. Whether a free dataset is a "commercial offer" is interpretation (I).
+4. "reach to API/PDF unverified": now checked. V: API docs, response headers and three sample payloads carry no terms (C1-C3); the Regulamin and Zarządzenie 3/2006 define SIS without naming the API (D1, D4, D5); the Sejm site footer lists the API and SIS as separate links (C4). I: placement of committee transcripts and interpellations in SIS by the Regulamin (D6, D7) is the only textual bridge.
+5. Missing from the statement: enforceability. ODA art. 15 ust. 1 gives a closed list of conditions that does not include a commercial-use bar; arts. 14 ust. 1, 16, 17 point the same way; art. 15 ust. 2 allows commercial-use limits only for museums, libraries, archives (L5-L9, V text; conclusion I). Sui generis database rights are the only plausible hook and only if the Kancelaria holds them (L6, L12); the Database Act voids contract terms that bar insubstantial parts (art. 7 ust. 4). Nothing read says the Kancelaria is a database producer.
+
+Replacement wording (suggested):
+
+"The Sejm's 'System Informacyjny Sejmu' page (www.sejm.gov.pl; wording in force since early 2016, absent on 2015-11-21) asks for a source note ('Źródłem tym jest System Informacyjny Sejmu') on material taken from www.sejm.gov.pl and bars SIS material from being an element of a commercial offer of any agency, TV station, portal or other entity. SIS is defined as the www.sejm.gov.pl information system (Regulamin Sejmu art. 201a; Zarządzenie Szefa KS 3/2006 § 1). No terms were found on api.sejm.gov.pl, in its documentation or in sampled API payloads, and no source states that API output is part of SIS. The Regulamin does place committee transcripts and interpellations in SIS, so the clause may be read to cover the same documents (interpretation). Its enforceability against reuse is doubtful under Open Data Act arts. 14-17 (the conditions an obliged body may set are listed in art. 15 ust. 1 and do not include a commercial-use bar), but is not settled by any authority read. The datasets' NOTICEs credit 'Kancelaria Sejmu RP, api.sejm.gov.pl' and do not use the SIS wording."
+
+Also for the registry cards: the committee card and NOTICE cite ODA arts. 15 as a basis. Art. 15 is the article that limits conditions; citing it does not support reuse unless the cited conditions (source, processing, liability, personal data) are actually met (I). The `sejm_api`/`sejm_interpellations` entries in `sources.py` cite arts. 2(12), 5, 14, 17 and omit 15/16.
+
+---
+
+## 4. Not checked
+
+- Live state of the Sejm10 terms page on the fetch dates (September 2026): live page returns the human-verification page; latest readable copy is Wayback 2026-06-03; the 2026-09-07 and 2025-01-17 copies are challenge pages. Earliest Sejm10 capture not found (CDX unavailable, Archive "Temporarily Offline").
+- Pre-2015 terms: Sejm7 and Sejm6 copyright pages had no CDX captures. Terms for content published before Nov 2015 (PPC and terms 7-8 in `sejm_api_2011_2022`) are unknown.
+- Exact introduction date and issuing act of the three-clause text (inside 2015-11-21..2016-02-04); no Wayback captures between those dates.
+- Kancelaria Sejmu BIP "Ponowne wykorzystywanie" page (ODA art. 11 ust. 1): bip.sejm.gov.pl did not resolve/connect; Wayback availability API answered 429 for seven URLs (one back-off, still 429), recorded as not checked. A BIP-published conditions page would be an ODA art. 12 offer and could change L10-L11. Also BIP.gov.pl subject page failed SSL verification in Python (not retried).
+- orka/orka2 terms and PDFs (bot block, not bypassed). A Marszałek Sejmu order cited for ZMS 1/2008 (nr 11/2015) was unreadable (403) and concerns building access.
+- Any other implementing act for SIS or iTV Sejm (Marszałek's zarządzenia on the website, iTV Sejm broadcasting rules).
+- Whether PiotrSty's `sejm_api*` text came from stenogram HTML, video-linked pages or another endpoint (README does not say; `/videos/{date}` was not fetched). Release and crawl dates of the HF release.
+- Whether API committee PDFs and interpellation bodies are the same documents as those on www.sejm.gov.pl (not compared).
+- IPI PAN crawl dates and the share of PPC text published after 2016-02; the ParlaCLARIN 2022 paper checked only by keyword (no channel named).
+- Case law, commentary or UOKiK/UODO/ministry guidance on (a) website-notice licence terms for public sector information, (b) commercial-use bars under ODA art. 15, (c) the Kancelaria Sejmu as a database producer. None searched beyond the statute texts.
+- Whether the Kancelaria Sejmu is formally an obliged body under ODA art. 3 (inference only, L3), and whether the SIS content is "informacja sektora publicznego" in the narrow sense for each document class.
+- dane.gov.pl: only API keyword searches; no manual browse.
+- Written MP questions (`writtenQuestions`) placement in SIS rests on art. 195 ust. 2 cross-reference; the fetcher's coverage of "pytania w sprawach bieżących" (art. 196, oral) not examined.