Continuity and Open Threads
This chapter collects every point on which the volumes of the series disagree about Nordvik Region, use one identifier twice, date one event twice, speak of different things under one name, or leave a plan without an outcome. It draws on the Open threads subsections of Chapters 3, 5 to 7 and 8 to 20, on the clashes recorded in Chapters 21 to 24, and on section 8 of each extract of the dossier, and it covers the fourteen volumes and the two companions Landscape and Tools (2, Nordqvist, 2026) (3, Nordqvist, 2026) (4, Nordqvist, 2026) (5, Nordqvist, 2026) (14, Nordqvist, 2026). The findings are for the author. The dossier does not choose silently: for each thread it states which volumes and chapters hold each side, the value the dossier works with while the thread is open, and the resolution it recommends. A recommended resolution is a proposal. A change to a volume is made in that volume's repository, as docs/UPDATING.md says, and a thread that the author closes is moved from this chapter to a note in the chapter concerned.
The register has 90 entries, numbered OT-NR-01 to OT-NR-96 with the numbers 06, 22, 23, 38, 50 and 73 withdrawn after review, in five kinds. A contradiction is two values or rules for one thing. An identifier collision is one identifier for two things, or two identifiers for one. A date clash is two dates for one event, or an order that cannot hold. A scope mismatch is a statement that, read in the Region's context, describes something the profile does not hold, or one name for two different scopes. A missing outcome is a decision, test, review or plan that a volume opens and no volume closes. Within each kind the entries run from the most consequential to the least. An entry is consequential when a reader who acted on the wrong side could set a control at the wrong level, accept a claim that has lapsed, or cite a record that is not the one meant.

Figure 27.1: The five kinds of open thread, who decides each and where the decision is recorded. Every kind ends with the author, because the organisation is invented and no external source can establish a fact about it.
Where the volumes label a 2027 to 2032 event as illustrative, the thread inherits the label. The book is written in November 2026, so only a small part of the Region's history is established (Chapter 21) and most of the dates below belong to the modelled future. Chapter 23 gives the figures behind several of the threads and Chapter 22 the records.
Contradictions
A contradiction is the most serious kind because the two values often govern a control. The first entries are the ones where a level, a limit or a time decides what a system may do or how fast it must be stopped.
| ID | Thread | Held by | Working value used by this dossier | Recommended resolution |
|---|---|---|---|---|
| OT-NR-01 | Stop of triage ranking: 60 seconds to date order, or within five minutes | 60 seconds: AI Governance, Chapter 10. Five minutes: AI Platform and Security, Chapters 5 and 24 | Both stated; Chapters 3 and 23 mark it open | Fix one figure for the stop and name the other as the time to confirm; the stop right is the most carefully assigned right in the series and its time should not have two values |
| OT-NR-02 | Level of the builders' assistant: Mode 2 and Mode 3 with ceiling Level 3 Bounded, or generate, refactor and test held at Level 1 with Mode 3 for the platform team only | Vibe to Dev (permit-by-tier.yaml, lane manifest; MND-NR-BLD-2031-01). AI Developer (DLG-NR-DEV-01; Chapter 32) |
Both stated; the builder levels are those of the builder volume, the developer ceiling is that of the developer volume | Decide whether the builders' lane is an exception to DLG-NR-DEV-01 and record the exception as an AIDR, or lower the lane ceiling |
| OT-NR-03 | Level of the platform team's agentic drafting: Level 2 from AIDR-1116 (2030-11-18), while the generate ceiling in DLG-NR-DEV-01 is Level 1 |
AI Developer, Chapter 32 and the profile | Level 2 from the date of the record; registration in the mandate is a first-quarter step | Amend DLG-NR-DEV-01 or MND-NR-DEV-01 so that the mandate and the ceiling agree, and say which holds when they differ |
| OT-NR-04 | Charter and manifest of the security architect: MND-NR-01 v1.0.0 and v2.0.0 allow Level 2 Supervised edits, the role charter caps at Level 1 Assisted |
The Transformed Security Architect, Chapters 25 and 26 | Charter governs; the volumes say the gap was never exercised | State in Chapter 25 whether the architect wrote every change from January because of the gap or in spite of it |
| OT-NR-05 | Ceiling of the ambient AI NR-AMB: Level 4 or Level 3 Bounded |
Level 4: the Landscape model. Level 3 Bounded: AI Red Team, Chapter 8 | Open; Landscape leaves the canonical value open | Name the canonical value; Level 4 for an ambient mode contradicts the series' own ceiling for unowned tools |
| OT-NR-07 | Level of the rostering system: Level 2 Supervised under DLG-NR-ROS-01, or Level 3 Bounded for M1 Direction in DRF-NR-01; open shifts at Level 3 from 2027-01-20 or 2027-03-11 |
Tools (Chapter 22) against DRF-NR-01 in AI Boss; Landscape model (2027-03-11) |
DLG-NR-ROS-01 v1.0.0 and 2027-01-20 as the record |
Align the Tools statement with the delegation profile and give the Landscape date as the date of the attestation, not of the change |
| OT-NR-08 | Block on rest breaches in DLG-NR-ROS-01: Level 3 Bounded in v1.0.0, or Level 1 Assisted in December 2026 and Level 3 only after CNT-NR-0001 (November 2027) |
AI Boss, Chapter 4 against the profile and Table 30.1 | The profile and Table 30.1 | Correct Chapter 4 or mark its v1.0.0 as the proposal |
| OT-NR-09 | Level 3 in operations: scripted runbooks only (profile); only the automatic rollback of a gateway route (Chapter 13); no Level 3 before stage N4 (Chapters 4 and 30) | AI Operations | Chapter 13: the rollback of a gateway route is the one Level 3 class today | State in the profile whether a Level 3 class exists and name it |
| OT-NR-10 | Modes of the security AI: Modes 0, 1 and 4, or also Mode 3 (platform team, Level 1); offensive ceiling Level 2 in both chapters | AI Red Team, Chapters 5 and 27 | Chapter 27 table | Add Mode 3 to Chapter 5 |
| OT-NR-11 | Tier-1 patch target: 48 hours to verified protection, or 24 hours to mitigation and 72 hours to fix (DPD-NR-001) |
AI Threat, Chapter 18. AI Red Team, Chapter 11 | Both stated; the volumes do not name each other's figure | Make 48 hours the verification target and name the two others as steps toward it, or the reverse |
| OT-NR-12 | Waiver length: at most 90 days (policy v2.1), or up to 180 days for a high-risk system (escalation matrix) | AI Governance, Chapter 13 | Both stated | Align the matrix with the policy, or write the 180 days as an exception for high-risk systems |
| OT-NR-13 | Log retention: 183 days, or "six months" (STD-AI-002); other records give 12 months, at least six for high risk, and 90 days |
AI Governance; the volumes of the series; Chapter 23 | Differs by system and by purpose; STD-AI-002 at 183 days |
Replace "six months" by 183 days and state which retention applies to which record |
| OT-NR-14 | Model version of benefits triage: 3.2.1 in the inventory, 3.2.0 in AIDR-0007 |
AI Governance, Chapters 9 and the AIDR | 3.2.0 for the decision, 3.2.1 for the running version | State that the AIDR records the version approved and the inventory the version running |
| OT-NR-15 | AI Act role of the staff assistant: provider (NR-UC-020) or deployer (UCD-2026-022) |
AI Governance, Chapters 8 and 15 | Deployer for a bought product, recorded as a thread in Chapter 5 | Correct NR-UC-020, or explain what the Region provides that makes it a provider |
| OT-NR-16 | Roles in a registry example: ai_act.roles: [deployer] for ranker 4.0.0 although the Region is provider of the explanation service |
AI Platform and Security, Chapter 21 and the profile | Provider and deployer of the explanation service; deployer of the ranker | Amend the example to list both roles |
| OT-NR-17 | What the caseworker sees: a band and three reasons (Chapters 10 and 25), an urgency_score (YAML of Chapter 10), or no raw score (Chapter 17) |
AI Governance | Band and three reasons; no raw score | Remove urgency_score from the YAML or label it as an internal field |
| OT-NR-18 | Whether the benefits triage and the living-support triage are one system or two | AI Governance, Chapter 10 (FRIA-2026-002) against Chapter 17 (NR-ALG-2026-003, lsb-triage v2.3.0); the profile says one claims ranker |
One system with two records (Chapter 5) | Decide, and give the system one NR-ALG identifier, one FRIA and one register entry |
| OT-NR-19 | Plan measures: four (PLAN-NR-SA-2027 in Chapter 26) or six (profile); five (N1 to N5, Chapter 32) or seven (N1 to N7, Appendix A and profile) for PLAN-NR-MGT-2027 |
The Transformed Security Architect; AI Boss | The four-measure and seven-measure versions of the chapters, as each volume's own chapter states | Bring the profiles and the chapters to the same number of measures |
| OT-NR-20 | Coverage claim: seven of nine had something behind them, or five tests with three failed; the ledger says five of nine with a result, two asserted, two untested | AI Purple, Chapters 1, 5 and 17 | The ledger: five, two, two | Correct Chapters 1 and 5 |
| OT-NR-21 | NIS 2 position: depends on national transposition, or the Cybersecurity Act applies (AIDR-0923, "essential operator") |
AI Operations, Chapters 2 and 3 against Chapters 23 and 24 | The Cybersecurity Act and the essential operator | Update the profile and Chapters 2 and 3 |
| OT-NR-24 | Size of the IT and Digitalisation Unit: about 85 with two vacant, or 82 as the five parts add up | AI Operations, Chapter 23 of this dossier | Both stated | Reconcile the parts or the total |
| OT-NR-25 | Volume of benefits triage: 40,000 applications a month (AI Governance, AI Platform and Security), 40,000 claims a month (AI Threat), about 31,000 a year (living-support summary); AI Environment also has about 40,000 claims ranked a month; Landscape no longer says that its own profile lacks the figure | AI Governance; AI Threat; AI Environment; Landscape | About 40,000 a month; see Chapter 26 | Define "application" and "claim", and reconcile with the 31,000 |
| OT-NR-26 | On-call load of the security analysts: one week in six at the start (Chapter 21) and one week in five after AIDR-0734 (2027-10-12); technical on-call in operations is a different rota, one week in eleven (AI Operations) |
AI Blue; AI Operations | A change over time, not a clash: one in six, then one in five; the operations rota is separate | Say in Chapter 26 which rota each figure concerns |
| OT-NR-27 | Local detection rules: "few" in the profile, twelve in Chapter 26 | AI Blue | Twelve | Change the profile |
| OT-NR-28 | Coaching of unit managers: ceiling 30 and floor 5 hours (SPN-NR-2029-01); target 28, ceiling 35, floor 4 hours (Chapter 16); interim three hours a week from 2030-01-08; three protected hours a month for new managers |
AI Leader, Chapters 11, 16 and 27 | Three measures coexist, each in its own record | Declare which supersedes which |
| OT-NR-29 | Practice before DLG-NR-EXE-01: Level 0 in most classes (Chapter 4) or Level 1 everywhere "a ceiling, not a licence" (profile) |
AI Leader | The profile as the ceiling and Chapter 4 as the practice | Say that the profile gives a ceiling and the chapter the practice |
| OT-NR-30 | Contestation reply: 14 days (AIDR-0624), 10 working days (FRIA-2026-007), 30 days (first schedule) |
AI Boss, Chapter 27 | Resolved by AIDR-0631: 10 working days, at most 14 calendar days |
Close; note the supersession in the chapter |
| OT-NR-31 | Dashboard classification: open in the profile, high-risk in AIDR-0609, which makes a small digitalisation unit a provider |
AI Boss, profile and AIDR | AIDR-0609 |
Say in the profile that the question is decided |
| OT-NR-32 | Patching and supplier amendment: NR-2025-044 is amended so that out-of-hours detections can isolate a service (AI Threat), the amendment took two procurement cycles (AI Red Team), and MS-2 is unsigned until 2028-03-31 (AI Blue) |
The three volumes | MS-1 existing, MS-2 pending |
State the amendment history once |
| OT-NR-33 | Negotiation with the unions before NR-SEC-001: stated in AI Threat, Chapter 13; absent from AI Red Team |
AI Threat; AI Red Team | Stated by AI Threat | Add the step to the profile |
| OT-NR-34 | Embedded workplace assistant pilot: class "accept and monitor" with a non-EU supplier (AI Threat), Mode 2 (profile), "pilot outside architecture work" (The Transformed Security Architect) | Three volumes | Mode 2 pilot, non-EU supplier recorded | Name the supplier and its jurisdiction in the profile |
| OT-NR-35 | Hosting and operation of benefits triage: operated by the Region on its own premises (operated_by: self, Landscape), or by TriageSoft in the Region's data centre (profile, AI Operations) |
Landscape; AI Platform and Security; AI Operations | Supplier-run in the Region's data centre | Correct the model |
| OT-NR-36 | Supplier of the recruitment pilot: non-EU supplier TalentLens for Nordvik (Landscape model), a module of an applicant tracking system for Nordvik and TalentLens for Hansa Mutual (AI Boss) | Landscape; AI Boss | AI Boss | Correct the model; the HRIS supplier (PeopleCore) is defined and no unit refers to it |
| OT-NR-37 | Owner of THR-NR-04: the CISO (Appendix A) or the Head of Benefits Unit, who owns HPR-NR-2026-01 and the system |
AI Threat, Chapter 11 | Both stated | Align the owner of a threat with the owner of the system it concerns |
| OT-NR-39 | Who adopted permit-by-tier.yaml: the digital steering group (profile) or the Head of IT and Digitalisation as owner (file, 180-day review) |
Vibe to Dev | Owner is the Head of IT and Digitalisation | Record who decided and who owns |
| OT-NR-40 | Stop budgets: 10, 14 and 15 minutes (AI Blue), 5 minutes and 120 seconds (AI Purple), 60 seconds | AI Blue; AI Purple; AI Governance | Different stops, not conflicting (Chapter 23) | Add a table of stops to a volume |
| OT-NR-82 | Energy per resolved enquiry of the open-weight route: 16.3 Wh (FPR-NR-0001 revision 1, 0.52 g) or 30.5 Wh (revision 2, 0.98 g) at the same facility boundary; 31.9 Wh in the reading of 2032-09-01; the register carried revision 1 at the snapshot of 2032-12-13; the three inputs that differ are draw (0.30 or 0.52 kW), PUE (1.15 or 1.21) and pool share (0.40 or 0.41) |
AI Environment, Chapters 6, 8 and 24 | Revision 2 as the later record, stated with its range; no independent check exists | Check draw and pool share independently before any statement or slack |
| OT-NR-83 | Boundary of the budget: EBG-NR-0001 is 4,450 kWh a month from 27.7 Wh (2026 factor, GPU servers and host, before the facility markup, times 1.1); revision 2 gives 25.2 Wh at that boundary and 30.5 Wh at the facility boundary (4,458 kWh); whether a use is "covered" depends on the boundary chosen |
AI Environment, Chapters 10, 15 and 19 | Three values stated; the budget as set | Name one boundary for budgets and for footprints |
| OT-NR-84 | Reserved accelerators: 16 reserved and up to 8 on demand in the capacity plan of 2027Q1; 24 and up to 8 in September 2032, with no volume between them recording the growth | AI Platform and Security, Chapter 22; AI Environment, profile | Different dates; both stated | Record the step from 16 to 24 and its decision |
| OT-NR-85 | The carbon figure of October 2026: 2,266 kWh at PUE 1.15, intensity 32 g, embodied 62,400 g, 71,000 cases, 1.90 g (AI Platform and Security), against 1,970 kWh of IT energy and a first treatment that does not say whether the facility overhead is included (AI Environment, Chapter 7); 1,970 kWh times 1.15 is 2,266 kWh | AI Platform and Security, Chapter 22; AI Environment, Chapter 7 | The two read as one figure at two boundaries | Say which boundary each volume uses |
| OT-NR-86 | Volume of the citizen-service assistant: 60,000 conversations a month (CAP-NR-AI-001, 2028), 71,000 resolved enquiries a month (October 2026) and about 168,000 (September 2032, a factor of 2.4) |
AI Operations, Chapter 15; AI Platform and Security; AI Environment | Different measures and dates; not reconciled | Define the unit and give one series |
| OT-NR-87 | The builder track of the footprint: FPR-NR-0006 and BEN-NR-0006 are attached to the builder workspace assistant (AIDR-1200, about 40 builders, about 9,000 sessions a month), while the 41 calls in August 2031 against 17 in January 2032 are the measure of one build, service-point-check, in a single uncontrolled month; Vibe to Dev has 140 FormFlow builders and 60 apps in 2030 |
AI Environment, profile and Chapter 4; Vibe to Dev; AI Developer | Two measures of two things; the count of builders differs by date | Say which build and which assistant each figure concerns |
| OT-NR-88 | The staff assistant: the adoption claim ADP-NR-2029-01 (61 per cent weekly use, recount AIDR-1020 with no recorded result) stands unchanged in AI Environment, which writes no adoption figure and no benefit record with a result; owner Henrik Dahl in AI Operations and AI Leader, use owner Gunilla Roos in AI Environment |
AI Leader; AI Environment | Open; the difference of owner is one of role | Record the recount and name the owner of the service and of the use |
The contradictions that matter most are OT-NR-01 to OT-NR-04. Each concerns a limit that can be tested: a stop time, a ceiling, a manifest. Sources: (2, Nordqvist, 2026) (3, Nordqvist, 2026) (4, Nordqvist, 2026) (5, Nordqvist, 2026) (6, Nordqvist, 2026) (7, Nordqvist, 2026) (8, Nordqvist, 2026) (9, Nordqvist, 2026) (10, Nordqvist, 2026) (11, Nordqvist, 2026) (12, Nordqvist, 2026) (13, Nordqvist, 2026) (15, Nordqvist, 2026) (16, Nordqvist, 2026).
Identifier collisions
An identifier that names two things breaks the lookup that Appendix A depends on. The first three entries are the ones that a reader citing a record would most likely get wrong.
| ID | Thread | Held by | Working value used by this dossier | Recommended resolution |
|---|---|---|---|---|
| OT-NR-41 | RST-NR-2031-0001 is the exit test of the code model (2031-02-03) and the builder's restore test (2031-12-05, 11 minutes, expires 2032-03-04) |
AI Developer, Chapters 26, 31 and 32. Vibe to Dev, Chapters 14 and 23 | Two records under one identifier (Chapters 19 and 22) | Renumber the later record, that of Vibe to Dev, to the next free RST-NR-2031-nnnn |
| OT-NR-42 | AIDR-1002 names both DLG-NR-EXE-01 (the leadership team's ceiling, 2029-10-01) and MND-NR-LDR-01; Tools also calls the regional director's profile DLG-NR-DIR-01, which AI Leader reviews on the same day |
Tools, Chapters 18 and 24; AI Leader | Both stated | Give each record its own AIDR |
| OT-NR-43 | NR-SEC-004 is a monthly restore drill (profile, Chapter 24) and a weekly chaos catalogue of twelve types (Chapters 25 and 27) of AI Red Team; the later volumes use it for the restore drill |
AI Red Team; The Transformed Security Architect; AI Operations | Restore drill and chaos catalogue are two programmes | Give the chaos catalogue its own identifier |
| OT-NR-44 | RVT-NR-2027-01 (a month without the claims ranker, AI Threat) and RVT-NR-ROS-2027-01 (rostering, 5 to 16 April 2027, AI Boss) are similar and different; the Landscape profile uses the second for both and its model has no record |
AI Threat; AI Boss; Landscape | Two tests of two systems | Give Landscape's model both identifiers |
| OT-NR-45 | NR-2026-031 is the staff assistant's hosting contract, the managed-inference tender (sovereignty profile) and the hosting provider's contract |
AI Platform and Security, Chapters 1, 13 and 24; AI Operations | The staff assistant's hosting contract | Give the tender its own number |
| OT-NR-46 | Registry identifiers: NR-ALG-2026-003 for living support in the profile and NR-ALG-2026-007 for the staff assistant; NR-UC-020 for the staff assistant against NR-ALG-2026-007 |
AI Governance, Chapters 8 and 15 | As the profile has them | Align the registry with the profile |
| OT-NR-47 | EU database identifier of benefits triage: EUDB-DEP-2026-004213 (Chapter 9) or EUDB-2026-NR-00412 (Chapter 17) |
AI Governance | EUDB-DEP-2026-004213 (Chapter 26) |
State that one of the two is wrong |
| OT-NR-48 | FRIA, DPIA and register records of benefits triage: FRIA-2026-002 or FRIA-2026-004; DPIA-2025-031 or -2026-007; NR-ALG-2026-003 |
AI Governance; Tools | Both stated | Align with OT-NR-18 |
| OT-NR-49 | Risk and finding names: NR-RISK-2027-014 or NR-RISK-AI-014 and -016; NR-FND-2027-014, FND-NR-2027-007 and FND-NR-2027-001-04 |
AI Red Team, Chapters 5, 13 and 15 | Chapter 5 names | One naming pattern for risks, one for findings |
| OT-NR-51 | ROLE-NR-MGR-01 is the unit manager's charter of 2027-09-01 (35 per cent of time protected, AI Boss) and the charter in force 2030-02-01 under AIDR-1016 (AI Leader) |
AI Boss; AI Leader | The second may revise the first | Add a version to the identifier |
| OT-NR-52 | AIDR-1102 is a decision in two parts, retrospective ambient completion and pilot continuation, cited differently in Chapters 1, 3, 4, 7, 16 and 20; the "rehearsal of 8 January" in Chapter 31 is not an AIDR |
AI Developer | Two parts | Split into two records |
| OT-NR-53 | Names in the series that resemble one another: Elin Berglund, Elin Berg, Sofia Berglund, Sigrid Berglund; Karin Lund, Frida Lund | AI Purple; AI Developer; AI Governance | Kept apart in Chapter 24 | Keep surnames distinct in later volumes |
| OT-NR-54 | Undefined or short-form identifiers: EVD-NR-2027-* cited and not defined; DLG-NR-LOOP-01 against DLG-NR-EMU-01; two forms of RoE-NR-2027-001; EXIT-NR-01 used twice; MND-NR-DEV-01 as a table row only; RISK-NR-0031 |
AI Blue; AI Purple; Tools | Listed in Appendix A as stated | Define or retire each |
| OT-NR-55 | Hanna Wikström is a product owner who builds prototypes (profile) and the builder of service-point-check (Vibe to Dev), three roles in one person |
AI Developer; Vibe to Dev | The builder volume | Say so in the profile |
| OT-NR-89 | FPR-NR-0007 is the next-use identifier in the profile of AI Environment and the model footprint sheet in its Chapter 5; the platform team's Mode 3 tool became FPR-NR-0008; EBG-NR-0002 (classifier) and EBG-NR-0003 (placement) were renumbered to avoid a collision; DLG-NR-ENV-01 became DLG-NR-ENV-0001 |
AI Environment, profile and Chapters 5, 15, 17 and 20 | The identifiers as the profile finally allocates them | Fix one allocation table |
| OT-NR-90 | AIDR-1300 to AIDR-1349: the rule "1300 plus chapter" is broken by second decisions from AIDR-1326; AIDR-1310, -1316 and -1323 are not listed although their chapters create records; AIDR-1318 is dated 2032-11-02 and cited for the district-heating study; HWL-NR-0001, TA-NR-ENV-01 and SDR-NR-0001 and -0003 carry no AIDR |
AI Environment, profile and Chapters 10, 16, 18 and 23 | Identifiers as listed | Complete the allocation table |
Sources for the collisions are the volumes cited in the table (12, Nordqvist, 2026) (13, Nordqvist, 2026) (16, Nordqvist, 2026) (7, Nordqvist, 2026) (11, Nordqvist, 2026).
Date clashes
A date clash is not always an error. Some are two chapters written at two dates, and the series says that each chapter's date governs. The entries are listed so that a reader who builds a timeline from the volumes does not meet them unprepared.
| ID | Thread | Held by | Working value used by this dossier | Recommended resolution |
|---|---|---|---|---|
| OT-NR-56 | AIDR-0211 accepted 2026-06-18, while the staff assistant was generally available from 2026-06-01 with about 3,200 users and AI Governance cites the record in April |
AI Platform and Security, Chapter 13; AI Governance | 2026-06-01 for use, 2026-06-18 for the record | Date the record in April or write that it ratifies |
| OT-NR-57 | Annex MS-2: effective 2027-07-01 (Chapter 16 draft), to be signed by 2028-03-31 (Chapters 26 and 27); status undecided at the end of the volume |
AI Blue | 2028-03-31, the later value | Say whether MS-2 was signed |
| OT-NR-58 | Restore test of 2031-12-05 against 2031-12-04 (expiry 2032-03-04 against 2032-03-03) | Vibe to Dev, Chapters 14 and 23 against Chapter 16 | 2031-12-05 | Correct Chapter 16 |
| OT-NR-59 | Step 9 gate: met 2031-11-12 and 2031-12-12 in the profile; rehearsal 2031-12-05 precedes v0.2.0 (12-08), yet the gate is met on 12-12 after the takeover test fails | Vibe to Dev, Chapter 13 and the profile | The later date | Give the first as the first attempt |
| OT-NR-60 | January measure (17 calls, measured_on: 2032-01-31) lies outside the four months of the volume; Chapter 23 says "not yet measured" |
Vibe to Dev, Chapters 23 to 25 | Not measured at the end of December | Move the figure to a later chapter |
| OT-NR-61 | Observer measurement: "not yet measured" (Chapter 30, dated 2029-03-31) against a seeded fault on 2029-03-19 (2 min 41 s, Chapter 29); retest planned 2029-06-30 against outcomes 2029-04-10 and 2029-06-05 | AI Operations, Chapters 29 and 30 | Chapter 29 | Date Chapter 30 earlier or revise it |
| OT-NR-62 | End of the evidence clock for the verified failover claim: 2029-12-02 (180 days from 2029-06-05) or 2029-12-27 before the rerun of 2029-12-15 | AI Operations, Chapters 29 and 30 | Both; no basis for choosing | Fix the clock and the length |
| OT-NR-63 | Failover test RST-NR-2027-002 of 2027-10-12: 5 hours 20 minutes against a target of 2 hours; AP-NR-OPS-03 gives 5:20, 2:00 and 3:40 without saying which belongs to which test |
AI Operations; Tools | 5 hours 20 minutes | Label each time with its test |
| OT-NR-64 | First out-of-hours stop drill: March 2027, 26 minutes against 15 (The Transformed Security Architect, Landscape); 2027-05-13, 3 min 50 s paging (AI Blue); "quarter two" (The Transformed Security Architect, Chapter 25) against the 2027 Q1 identifier | The three volumes; the factcheck of Landscape on its Chapter 13 (pointer to its Chapter 18, which gives only the other organisation's 17 and 5 minutes, while the profile points to Chapter 10) | Two drills of two stop paths | Say which drill each volume means |
| OT-NR-65 | Annex MS-1: existing at the start (AI Blue, AI Red Team), signed May 2027 (The Transformed Security Architect) |
Three volumes | Existing, with May 2027 as an amendment | State it once |
| OT-NR-66 | AIDR order: AIDR-0601 (2026-09-18) before AIDR-0600 (2026-11-16); AIDR-0802 (02-22) numbered before -0803 and -0804 (02-08); AIDR-0813 (02-05) against the MS-2 window of 02-07 to 02-20; AIDR-0713 (conclusion) lower than -0716 but late; gaps in AIDR-0807 to -0811 and others; AIDR-1001 and -0916 undated; AIDR-1202 unused; review date 2031-03-23 is a Sunday (held Friday 03-21) |
AI Boss; AI Purple; AI Blue; AI Operations; AI Leader; AI Developer; Vibe to Dev | Numbers are identifiers, not an order; Chapter 22 | Add dates where missing and renumber where a number precedes its own cause |
| OT-NR-67 | DLG-NR-SEC-01 v1.1.0 dated 2027-03-22, yet PBK-NR-ISOLATE-01 in force in January; MND-NR-01 review_by 2027-05-01 and v3.0.0 decided in May |
AI Blue, Chapter 26; The Transformed Security Architect | The dates as stated | Date the playbook before the profile or the other way round |
| OT-NR-68 | Start of benefits triage: 2026-01-01 (Landscape), living-support triage in use 2026-03-03 (AI Governance); AIDR-0007 and -0018 dated 2026-04-15 and 2027-02-10; 4.0.0 shadow 2027-01-29 |
AI Governance; AI Platform and Security | Both | Date the records at the time of the decision |
| OT-NR-69 | Motion in the council: September 2026 without a day, ATA-NR-2026-01 commissioned 2026-10-05; not mentioned by The Transformed Security Architect; HP-5 open to 2027-03-31 against "missing link now exists" in Chapter 33 |
AI Threat; The Transformed Security Architect | September 2026 | Give the date and close or reopen HP-5 |
| OT-NR-91 | The conditions of AIDR-1326 arrived on 2033-02-25, 2033-03-11 and 2033-04-28 (Chapter 27) against due dates of 2033-02-28, 2033-03-14 and 2033-04-29 (Chapter 28); AIDR-1327 is dated 2033-05-10, after all three; TA-NR-ENV-01 is issued on 2033-02-14 while AIDR-1318 of 2032-11-02 places its ten questions in the annex |
AI Environment, Chapters 18, 27 and 28 | Arrival dates as the record; the volume calls the timeline illustrative | Date each condition once |
| OT-NR-92 | Article 49: AI Boss expects registration under Article 49(3) from 2027-12-02 and also says that Article 49 is not deferred; Article 113, third paragraph, point (c), names Chapter III, Sections 1, 2 and 3 for 2 December 2027, and Article 49 is in Section 5; Landscape now words its glossary so that registration is not grouped with the deferred duties | AI Boss; Landscape; the consolidated AI Act | Dates for Articles 26 and 27 as stated; no separate date for Article 49 is drawn | State the date that applies to the Region's registration |
Scope mismatches
A scope mismatch arises when an example in a volume does not belong to the Region as the profile has it, or when one name covers two scopes. The first entry affects the whole dossier.
| ID | Thread | Held by | Working value used by this dossier | Recommended resolution |
|---|---|---|---|---|
| OT-NR-70 | The state: "a Nordic member state" in the profile; Sweden in AI Operations (Cybersecurity Act, MBL section 11), AI Leader (Kommunallag, MBL, Cybersecurity Act) and AI Boss (MBL, working environment law); named Sweden in AI Environment (Lag (2025:570)) | AI Operations; AI Leader; AI Boss; AI Environment | "A Nordic member state", with Sweden named in AI Environment and Swedish sources noted where a volume uses them | Name the state, or state that the Swedish sources are examples |
| OT-NR-71 | Examples that may not be the Region: "the social insurance authority introduced in earlier chapters" in the routing and incident examples of Chapters 21 and 22, and the alert group claims-routing-ai with a rule for a maintenance-agent the Region does not have |
AI Platform and Security | Treated as the Region's routing, with the agent rule outside the profile | Say whose examples they are and remove the agent rule |
| OT-NR-72 | Person-level traces in examples: the token example officer-30417 at idp.nordvik.example, and the model name model-risk/a.lind in a registry example, the only personal names in the volume's material on the Region |
AI Platform and Security, Chapter 12 and the registry example | Not used as people | Replace with role names |
| OT-NR-74 | Agent that the Region does not have: the indicator table of Chapter 30 cites a drill "of the Region's agent" on 2028-11-07; NR-OPS-002 and -003 enter in Chapter 16 and not in the profile |
AI Operations, Chapters 16 and 30 | The drill concerns the assistant or the gateway; the two identifiers as stated | Reword the indicator; add the two to the profile |
| OT-NR-75 | Titles and names of posts: Chief Digital Officer and Head of Digital Services (AI Governance, AI Platform and Security), head of digitalisation (Landscape), Head of IT and Digitalisation (AI Operations onward); regional director unnamed except in AI Leader (Elisabet Holmgren), regionchef in AI Boss; Head of Platform and Maja Lindqvist; Karin Lund as mentor (Chapters 17 and 26 of AI Blue) and not as mentor (Chapters 21 and 23) | AI Governance to AI Leader; AI Blue | Posts not assumed to be the same office unless a later volume says so (Chapters 3 and 24) | Give one title per office and say who holds it |
| OT-NR-81 | Size of the regional management group: ten members (profile), but the pilot of the embedded workplace assistant is extended to "the 18 members" of it | AI Leader, profile and Chapters 22 and 27 | Ten members in the profile; the 18 may include assistants or deputies | Give one count and say who the 18 are |
| OT-NR-76 | Hosting picture: own cloud in two halls (Tools), staff assistant at a European cloud provider (NR-2026-031), citizen service on reserved capacity |
Tools; AI Platform and Security | Different parts of the estate (Chapter 6) | Add one hosting map |
| OT-NR-93 | The body that decides environmental budgets: AI Environment convenes a regional AI board "by the AI governance lead of the earlier volumes"; the earlier volumes have tier 3, the AI governance board with the regional director | AI Environment; The Transformed Security Architect | The same body, until a volume says otherwise | Name one body |
| OT-NR-94 | The service desk: AI Operations has Sara Ekdahl as service desk lead; AI Environment has her as service owner of SVC-NR-005 and also a service desk manager, not named and outside the sponsor line, who owns BEN-NR-0001 |
AI Operations; AI Environment | Not assumed to be the same person | Say whether the manager is Sara Ekdahl or another post |
The scope threads rarely change a control. They matter because a reader who treats an example as the Region's description attributes to it an authority, an agent or a person it does not have. Sources: (3, Nordqvist, 2026) (10, Nordqvist, 2026) (8, Nordqvist, 2026) (11, Nordqvist, 2026).
Missing outcomes
A missing outcome is a plan or decision that a volume opens and no volume closes. These are not errors: the series ends at a date and the Region goes on. The recommended resolution is in each case to write the outcome or to declare it outside the series.
| ID | Thread | Held by | Working value used by this dossier | Recommended resolution |
|---|---|---|---|---|
| OT-NR-77 | Open items at the end of AI Developer: tender NR-2030-041; the lawyer's position on steward status (due 2031-03-31); the register reread of 2031-07-21; the archivist's position on prompts; the deputy for the register adapter (to be named by 2030-12-15); MBA-NR-2030-03 (prompt retention, consultants); break-glass accounts to be tested by 2031-09-30 |
AI Developer, Chapters 31 and 32 | No outcome | Write each outcome in the next volume |
| OT-NR-78 | Possible breach of MBL-NR-IT-2027-01 by extending the pilot without negotiation, discussed and not decided; the Understand test failed live in PIR-NR-2030-0006 and no other live test is recorded; pilot figures (31 per cent, 9 of 12, "a third faster") lack a measured delivery time and the usage claims expired in July 2030 and the value claim in October 2030 |
AI Developer, Chapters 3, 22 and 32 | Open; claims expired | Decide, test again and retest the claims |
| OT-NR-79 | Open items at the end of the year in AI Leader: the staff assistant not covered by MBA-NR-2029-02; no exit test on benefits triage; DRR-NR-001-LEG not reread (lapsed 2030-06-10, derived); the 61 per cent figure not recalculated (AIDR-1020, 2029-11-30); span and coaching not evaluated; the carer recount and the Strandvik pulse result missing; no reading date for the Region's duty register |
AI Leader, Chapters 26, 31 and 32 | Open | Give each a result or a date |
| OT-NR-80 | Reviews with no recorded outcome: the profile and briefing specification (2030-04-01 and -02), SPN-NR-2029-01 (2030-05-05), LDR-NR-2029-0004 (May), CMP-NR-2029-21 (2030-05-12), LDR-NR-2030-0026 (2030-09-01); in other volumes the status of MS-2, the renewal of the call-off of Elin Berglund (ends 2028-11-30), FND-NR-2028-003 (closes after the test of 05-02), T1059.001 as a declared gap with no finding, COV-NR-0005 failed without a finding, the time achieved in the first stop-path test of DRF-NR-01, HP-5 and the accelerator dependency THR-NR-07, the sandbox gap sandbox_fail_closed: missing of NR-UC-017, the capacity of the platform team and one receiving developer, and the Landscape author's notes |
AI Leader; AI Blue; AI Purple; AI Boss; AI Threat; AI Platform and Security; Vibe to Dev; Landscape | Open; the dossier invents none | Close or declare each as outside the series |
| OT-NR-95 | Open at the end of AI Environment (August 2033): the re-tender of NR-2026-031 (with a procurement lead time of 9 to 14 months); CLM-NR-0001 withdrawn, with republication conditional on an independent check of FPR-NR-0001 and a statement about the hosted route; the eligibility route, held by a cap of 25,000 and with no energy figure; the footprint of the benefits-triage ranking (needs the AI share of nordvik-dc and the supplier's retraining energy); the district-heating study; SDR-NR-0002; the time shift of placement; about EUR 65,000 a year for four empty accelerators until the contract gives an option to shrink; the rebound check of the staff assistant after 90 days (about 2033-08); no independent check of FPR-NR-0001 |
AI Environment, Chapters 18, 27 and 28 | Open; the volume expects the re-tender to stay open after August 2033 and the claim rows to stay mostly empty in the second year (expected) | Write each outcome in a later volume or declare it outside the series |
| OT-NR-96 | The pointer for the first out-of-hours stop drill in Landscape, Chapter 13, goes to its Chapter 18, which gives only the other organisation's 17 and 5 minutes; Landscape Chapter 3 states the Region's 26 and 15 minutes | Landscape (factcheck, ch_13) |
The profile's Chapter 10 reference | Correct the pointer |
Reading the register
The register is long, but three patterns carry most of it. The first is that identifiers are reused across volumes written one after the other, and the cause is the series order: each volume reuses the Region's record families and does not look back. This is the cause of OT-NR-41 and OT-NR-42. The second is that a profile and the chapters of a volume disagree, as in OT-NR-19 and OT-NR-27; the cure is to run the profile and the chapters against the same ledger. The third is that the later volumes open reviews whose outcome falls after the volume's own horizon, which is why Chapter 23 gives an evidence age for each claim.
The recommended order for the author is the order of the entries: first the limits (OT-NR-01 to OT-NR-12), then the identifiers that collide (OT-NR-41 to OT-NR-45), then the dates that govern a claim's evidence age (OT-NR-58, OT-NR-61 and OT-NR-62), and last the scope and outcome threads. scripts/check_ledger.py lists the identifiers that a volume names for the Region and the ledger does not hold, and a thread that the author closes is moved to a note in the chapter concerned (docs/UPDATING.md).
Summary
The discord about Nordvik Region in the series falls into five kinds: contradictions of limits, levels and facts, identifier collisions, date clashes, scope mismatches and missing outcomes. None of them is a failure of the series as a whole. They arise from the way the series is made, one volume after another, each with its own subject and its own chapter dates, and from the fact that the Region has one profile and fourteen sets of chapters. The most consequential are the stop time of the benefits-triage ranking, the ceilings of the builders' assistant and of the platform team's agentic drafting, the mandate and the charter of the security architect, and the reused identifier RST-NR-2031-0001. The dossier does not resolve any of them. It states the working value it uses and the resolution it recommends, and it leaves the decision to the author, because the Region is invented and only its author can say which side is true.
Sources
- Nordqvist, G. AI Governance: Governing Artificial Intelligence in European Enterprises and Public Authorities. (2, Nordqvist, 2026)
- Nordqvist, G. AI Platform and Security: Building and Securing AI Platforms in European Enterprises and Public Authorities. (3, Nordqvist, 2026)
- Nordqvist, G. AI Red Team: AI as an Active Part of Cyber Defence in European Enterprises and Public Authorities. (4, Nordqvist, 2026)
- Nordqvist, G. AI Threat: Are the Dystopias Coming True? Evidence, Risk and Response for European Enterprises and Public Authorities. (5, Nordqvist, 2026)
- Nordqvist, G. The Transformed Security Architect: Role, Craft and Career in the Age of AI in European Enterprises and Public Authorities. (6, Nordqvist, 2026)
- Nordqvist, G. AI Boss: When Artificial Intelligence Manages People in European Enterprises and Public Authorities. (7, Nordqvist, 2026)
- Nordqvist, G. AI Blue: How Artificial Intelligence Transforms the Blue Team in European Enterprises and Public Authorities. (8, Nordqvist, 2026)
- Nordqvist, G. AI Purple: How Red and Blue Learn Together When AI Joins the Loop in European Enterprises and Public Authorities. (9, Nordqvist, 2026)
- Nordqvist, G. AI Operations: How Artificial Intelligence Transforms IT Operations in European Enterprises and Public Authorities. (10, Nordqvist, 2026)
- Nordqvist, G. AI Leader: How Leadership Roles Change When AI Is Used Broadly in European Enterprises and Public Authorities. (11, Nordqvist, 2026)
- Nordqvist, G. AI Developer: How the Developer Role Changes When AI Is Used Broadly in European Enterprises and Public Authorities. (12, Nordqvist, 2026)
- Nordqvist, G. Vibe to Dev: From Vibe Coding to Professional Software, Step by Step, for Builders Who Are Not Developers, in European Enterprises and Public Authorities. (13, Nordqvist, 2026)
- Nordqvist, G. AI Environment: Designing AI for the Greatest Benefit and the Least Environmental Harm in European Enterprises and Public Authorities. (14, Nordqvist, 2026)
- Nordqvist, G. Landscape: A Fact-Based Picture of Architecture and AI at Work, for Leaders, Architects and Citizens. (15, Nordqvist, 2026)
- Nordqvist, G. Tools: A Reference to the Templates, Software and Methods of The Transformed Architecture. (16, Nordqvist, 2026)