{
  "decision_id": "dvr_2Hd2UH3Djg1JU9Gv",
  "reason_codes": [],
  "policy_epoch": "pe_sandbox_v1",
  "decided_at": "2026-07-10T00:40:37.358Z",
  "korero_sha256": "5996621f78565c352e8ac45a5ef2a11b5b5e9a5340913a6479cfbea293ab7c19",
  "tenant_did": "did:web:devtest.mysovereignty.digital",
  "record_type": "deliberation_record",
  "record_frozen": true,
  "context": {
    "tags": []
  },
  "korero": {
    "proposal": {
      "title": "Strategic Posture",
      "description": "Deliberation strand strategic"
    },
    "participants": [
      {
        "ref": "p1",
        "role_label": "member",
        "did": null
      },
      {
        "ref": "p2",
        "role_label": "facilitator",
        "did": null
      },
      {
        "ref": "p3",
        "role_label": "member",
        "did": null
      },
      {
        "ref": "p4",
        "role_label": "member",
        "did": null
      }
    ],
    "entries": [
      {
        "seq": 1,
        "at": "2026-07-10T00:37:59.329Z",
        "participant_ref": "p2",
        "kind": "statement",
        "body": "This strand sets pace and posture. The evidence base is blunt: New Zealand was the last OECD nation to publish a national AI strategy, and the strategy we got — Investing with Confidence, July 2025 — is an adoption strategy, not a capability one [S-10]. The framework under test already commits to lifting R&D investment and backing proven accelerator programmes. The comparator set shows small nations buying capability through funds and partnerships: Canada's compute access fund up to CA$300M, Singapore's compute initiative up to S$150M [S-1]. Our question: adopt, build, or pool — and in what mix, gated by what evidence? — Heath Calder (chair)",
        "refs": []
      },
      {
        "seq": 2,
        "at": "2026-07-10T00:37:59.355Z",
        "participant_ref": "p2",
        "kind": "statement",
        "body": "Let me open with the productivity economist's caution. The case for AI adoption in a low-productivity economy is strong and the official strategy, whatever its ambition ceiling, points at the right sectors — health, agriculture, education [S-10]. The case for building capability is real but thinner in evidence: every dollar of build is a dollar not spent on diffusion, and diffusion is where measured productivity gains actually show up. So my prior is fast-follower with selective build, and I want every build proposal in this strand to carry an evidence gate — a named condition under which we stop. This working group's register is evidence-based policy; that has to mean we are willing to be wrong in public about our own enthusiasms.",
        "refs": []
      },
      {
        "seq": 3,
        "at": "2026-07-10T00:37:59.372Z",
        "participant_ref": "p4",
        "kind": "statement",
        "body": "Heath, adoption-only has a structural problem your framing skips: when everything we adopt is built offshore, the margin, the capability, and the exit options all sit offshore too. I have built and sold SaaS from here — New Zealand firms can win globally in niches where we hold the domain data and the domain problem: agritech, health workflows, te reo language technology. That is not frontier fantasy, it is an evidence-based investment agenda made concrete: lifting R&D investment and doubling down on proven accelerator programmes. The strategy's own admission — a shortage of AI expertise, capability framed as adopting rather than building [S-10] — is a choice, not a law of nature. I want this strand to say: niche-builder, deliberately, in two or three domains where we have durable advantage.",
        "refs": []
      },
      {
        "seq": 4,
        "at": "2026-07-10T00:37:59.390Z",
        "participant_ref": "p3",
        "kind": "statement",
        "body": "From the trade side, both of you are underweighting the partnership column. Small advanced economies do not choose between adopt and build; they pool. Canada put up to CA$300M into a compute access fund and runs a pre-approved AI supplier list; Singapore's Enterprise Compute Initiative is up to S$150M in credits and consultancy; Ireland and Estonia route capability through skills programmes and public-sector platforms rather than sovereign labs [S-1]. And the quiet lever is standards: New Zealand has a seat at ISO/IEC JTC 1/SC 42 if we resource it [S-1]. My caution runs the other way from Tom's: whatever posture we pick must not diverge from the regulatory regimes of our trading partners, because divergence is a non-tariff barrier we impose on our own exporters. Posture is also a foreign-policy instrument.",
        "refs": []
      },
      {
        "seq": 5,
        "at": "2026-07-10T00:37:59.403Z",
        "participant_ref": "p2",
        "kind": "statement",
        "body": "GAP acknowledgement — G-3. Before we vote on posture, the record should hold what does not exist: New Zealand has no sovereign frontier-capability programme of any kind — no national lab, no foundation-model effort, no frontier compute [S-10][G-3]. The strategy explicitly frames national capability as adopting AI, and its infrastructure posture centres on foreign-owned hyperscale cloud [S-10]. I am not posting this to end Tom's argument — niche-building does not require frontier capability — but to bound it. Any option this strand ranks must be honest that the starting stock is zero, so year-one deliverables are institutional (fund design, domain selection, partnership terms), not technical. And my evidence gate stands: each funded domain gets a named review point with a stop condition.",
        "refs": []
      },
      {
        "seq": 6,
        "at": "2026-07-10T00:37:59.417Z",
        "participant_ref": "p4",
        "kind": "statement",
        "body": "Accepted on both counts, and the comparator set actually sharpens my case rather than undercutting it. Look at Denmark: sovereign public-sector AI running on home-or-EU-jurisdiction compute, plus a regulatory sandbox [S-1] — that is a niche-builder posture from a nation our size, and nobody there is pretending to race the frontier. The distinction I want in the draft: frontier-chasing is out (Heath's G-3 bound is real [G-3]), but domain capability — models and products fitted to problems we own, accountable to people here [S-1] — is exactly what a capability fund buys. Fund it through instruments that lift R&D investment, gate it with Heath's stop conditions, and aim it at two or three domains, not a scatter.",
        "refs": []
      },
      {
        "seq": 7,
        "at": "2026-07-10T00:37:59.439Z",
        "participant_ref": "p3",
        "kind": "statement",
        "body": "Then the bridge is straightforward: niche-builder and partnership are the same posture seen from different desks. The fund's compute-access line is the pooling instrument — we buy access alongside partners rather than building alone, which is the cross-cutting lesson of the comparator table: home-or-allied-jurisdiction compute is the decisive lever, not ownership of buildings [S-1]. The domain-capability line is Tom's, evidence-gated per Heath. And the standards seat [S-1] plus alignment discipline is the wrapper that keeps our exporters inside our partners' regimes. I can rank a blended option first if it names all three components rather than burying the international one in an annex — that is my condition, and it is substantive, not cosmetic.",
        "refs": []
      },
      {
        "seq": 8,
        "at": "2026-07-10T00:37:59.456Z",
        "participant_ref": "p2",
        "kind": "statement",
        "body": "Synthesis, then to decision. Posture: evidence-gated niche-builder within partnership. Concretely — (1) baseline adoption push continues; the official strategy's sector focus is right even where its ambition is low [S-10]. (2) A capability fund on the Canada/Singapore pattern [S-1], sized in the cross-party proposal's indicative envelope, funds domain capability in two or three named domains, each with a review point and a stop condition. (3) The fund's compute line buys pooled access with partners, not sovereign scale [S-1][G-3]. (4) Standards engagement at SC 42 and regime alignment with trading partners as standing posture [S-1]. Funded consistently with lifting R&D investment. Tom moved me off adoption-first; Ana bound us to alignment. Recording both.",
        "refs": []
      }
    ],
    "ai_inputs": [],
    "engagements": []
  },
  "korero_segments": [
    {
      "seq": 1,
      "submitted_at": "2026-07-10T00:40:37.357Z",
      "segment_sha256": "e9c188497ef63c2a399cfd4de8220a3a4697f5100458b92df4fdf47b570c71b5",
      "entry_range": {
        "from": 1,
        "to": 0
      }
    },
    {
      "seq": 2,
      "submitted_at": "2026-07-10T00:40:37.376Z",
      "segment_sha256": "88d19c34108929485f8becdbbe225bcba0e9d9605bc51af22c4c99d4b468021e",
      "entry_range": {
        "from": 1,
        "to": 8
      }
    }
  ],
  "receipts": [
    {
      "receipt_version": 2,
      "receipt_id": "gr_ZwqmTAbcdd5H4Xxd",
      "decision_id": "dvr_2Hd2UH3Djg1JU9Gv",
      "tenant_did": "did:web:devtest.mysovereignty.digital",
      "request_sha256": "e782f1557268f3bba7f26d89e5b9976a9ffd66fcacbc7dd16b927b184601779b",
      "action_payload_sha256": null,
      "korero_sha256": "e9c188497ef63c2a399cfd4de8220a3a4697f5100458b92df4fdf47b570c71b5",
      "korero_segment_count": 1,
      "reason_codes": [],
      "policy_epoch": "pe_sandbox_v1",
      "decided_at": "2026-07-10T00:40:37.357Z",
      "algorithm": "Ed25519",
      "signer_did": "did:web:devtest.mysovereignty.digital#prf-b98f1a16593cead6271edd93e4b24888",
      "signature": "9291656e736d7207456f79774dcada898811b49d65e64b48bd159aa370e04fe6a2b52feadddff9ece200ee1f95287446f856bf9aea5124248200c2a61a87530f",
      "receipt_kind": "record_open",
      "integrity_mode": "tamper_evident",
      "signing_model": "tenant_key"
    },
    {
      "receipt_version": 2,
      "receipt_id": "gr_0ZtY7nLtr0gLW1CF",
      "decision_id": "dvr_2Hd2UH3Djg1JU9Gv",
      "tenant_did": "did:web:devtest.mysovereignty.digital",
      "request_sha256": null,
      "action_payload_sha256": null,
      "korero_sha256": "5996621f78565c352e8ac45a5ef2a11b5b5e9a5340913a6479cfbea293ab7c19",
      "korero_segment_count": 2,
      "decision": null,
      "reason_codes": [],
      "policy_epoch": "pe_sandbox_v1",
      "decided_at": "2026-07-10T00:40:37.376Z",
      "algorithm": "Ed25519",
      "signer_did": "did:web:devtest.mysovereignty.digital#prf-b98f1a16593cead6271edd93e4b24888",
      "signature": "03040052487c9104471f5ee090ce5954be7bdc937e739d69d553d7d816490554cc174f86622005a76e3993302c8ee985f67161e6f19bfc300fc03ab626a13501",
      "receipt_kind": "record_segment",
      "integrity_mode": "tamper_evident",
      "signing_model": "tenant_key"
    },
    {
      "receipt_version": 2,
      "receipt_id": "gr_QjYref3_5HYwvOl4",
      "decision_id": "dvr_2Hd2UH3Djg1JU9Gv",
      "tenant_did": "did:web:devtest.mysovereignty.digital",
      "request_sha256": null,
      "action_payload_sha256": null,
      "korero_sha256": "5996621f78565c352e8ac45a5ef2a11b5b5e9a5340913a6479cfbea293ab7c19",
      "korero_segment_count": 2,
      "decision": null,
      "reason_codes": [],
      "policy_epoch": "pe_sandbox_v1",
      "decided_at": "2026-07-10T00:42:11.235Z",
      "algorithm": "Ed25519",
      "signer_did": "did:web:devtest.mysovereignty.digital#prf-b98f1a16593cead6271edd93e4b24888",
      "signature": "d56c88334d1d72e6c0b8c608e55767e27485b68aff3e68b90b83f9e39d190061acee6598b3808595fe57b6e48671332d81c39b7a064165b59218a1c1cbb6ef01",
      "receipt_kind": "record_seal",
      "integrity_mode": "tamper_evident",
      "signing_model": "tenant_key"
    }
  ],
  "proof_chain": [
    {
      "boundary_crossed": "create",
      "policy_evaluated_by": "sovereignRecordPlugin.create",
      "decision": "allow",
      "caveats_added": [],
      "timestamp": "2026-07-10T00:40:37.361Z",
      "algorithm": "Ed25519",
      "signature": "13cd4193f42e4d87413389a448568cb265246fb98c47703ecd40ace982813a45976e550fc2ee91685ceededa81aa412adb470696b2df7116a3064e4ff18c300b",
      "signer_did": "did:web:devtest.mysovereignty.digital#prf-b98f1a16593cead6271edd93e4b24888"
    },
    {
      "boundary_crossed": "update",
      "policy_evaluated_by": "sovereignRecordPlugin.update",
      "decision": "allow",
      "caveats_added": [
        "korero",
        "korero.proposal",
        "korero.proposal.description",
        "korero_segments",
        "receipts",
        "korero_sha256",
        "metadata",
        "metadata.origin",
        "metadata.policy",
        "metadata.encryption"
      ],
      "timestamp": "2026-07-10T00:40:37.384Z",
      "algorithm": "Ed25519",
      "signature": "3beeabde59560803c5a53f1809c8fd379ffcee217050d61cc926d0ed1517a0aceaff4b9d6bf4591dc27d1799b758f9293d4fe94d31c9ea6a82d6ee7fe11b8c03",
      "signer_did": "did:web:devtest.mysovereignty.digital#prf-b98f1a16593cead6271edd93e4b24888"
    },
    {
      "boundary_crossed": "update",
      "policy_evaluated_by": "sovereignRecordPlugin.update",
      "decision": "allow",
      "caveats_added": [
        "record_frozen",
        "receipts"
      ],
      "timestamp": "2026-07-10T00:42:11.236Z",
      "algorithm": "Ed25519",
      "signature": "8069239953ed8cb56aaa8f8549b841feca54825672b67c64e16ceeeefb34a4e7f6a38f8bca8ac5a66e3f44f74c60af346fc0047e10f6ffe1720a48258f9d2d00",
      "signer_did": "did:web:devtest.mysovereignty.digital#prf-b98f1a16593cead6271edd93e4b24888"
    }
  ],
  "origin": {},
  "verification_keys": [
    {
      "id": "prf-b98f1a16593cead6271edd93e4b24888",
      "type": "Ed25519VerificationKey2020",
      "publicKeyMultibase": "z6MkoJz8eJCKmafbRaPYaheAzeRcZxETJxEszRj87YW2FgCd",
      "algorithm": "Ed25519",
      "created_at": "2026-04-22T18:55:57.179Z"
    }
  ],
  "retention": {
    "legal_hold": false,
    "retain_until": null
  },
  "verification": {}
}