The designed responses
Surfaces
Every response surface the library has built, rendered live. Browse by territory; each links to the full pattern.
T1 Provenance & Intent Telling where a submission came from, and whether it states a view the user holds or one an agent generated. A fluent submission no longer proves a user wrote it, yet the agency still has to act on it.
7 of 7 built T2 Delegation & Authorization Letting a user give an AI agent scoped, revocable authority an agency can check is genuine and current. Government's identity systems can prove who a person is, but not what they let an agent do for them.
13 of 13 built T3 Accountability & Audit Keeping what an AI agent did in a user's name visible and answerable, for one case and across a whole population. Mistakes tend to surface only after the damage is done, and one faulty process can repeat on everyone at once.
8 of 8 built T4 Trust Calibration Helping a user judge how far to rely on a government AI agent, and giving the agency a basis to support that judgment. A user can wrongly refuse an agent that would help, or lean on one a serious case needs a human for.
7 of 7 built T5 Rationing & Friction Forms, queues, and deadlines rationed services for years by demanding effort no one designed in. AI agents strip that effort away, forcing a deliberate choice about which friction only excluded people and which metered need.
7 of 7 built T6 Volume vs Breadth Signaling Reading the weight of opinion behind public submissions once agents file fluent, distinct-looking entries at scale. Volume and the look of variety can both be manufactured, so neither shows how widely a view is held.
8 of 8 built T7 Representation & Equity Keeping every user able to reach a government outcome, from a consultation to a benefit, license, or appeal, as private agents come between them and government. Left alone, outcomes start to track who has the better agent.
9 of 9 built T8 Verification & Certification Knowing a reliable civic tool from a confident-sounding wrong one, for users and the automated tools government runs itself. Building an authoritative-looking tool is nearly free; verifying it and standing behind it is not.
8 of 8 built T9 Cross-border & Sovereignty What changes when the AI model behind a service runs, or can be legally compelled, outside a user's own legal protection. It decides whose data law applies and who can reach the data, often without either side able to tell.
8 of 8 built
T1 Cryptographic content provenance Provenance Receipt Built T1 Verifiable credentials and decentralized identity Share Credentials Built T1 Self-attestation and disclosure Preparation Disclosure Built T1 Proof of personhood at submission Confirm Personhood Built T1 Structured intake with process metadata Guided Intake Built T1 Linking a person's input to the final text Composition Timeline Built T1 The attestation-verification gap Tiered Attestation Built T2 Fine-grained scope negotiation Negotiate Scope Built T2 Digital power of attorney Manage Authority Built T2 Consumer data rights consent flows Data-Access Consent Built T2 Consent receipts and records Delegation Record Built T2 Delegation registries Verify Delegation Built T2 Binding an agent to a verified identity Bind Identity Built T2 Step-up re-authorization for sensitive actions Step-Up Authorization Built T2 Nominated-agent authorization Nominate Agent Built T2 User-defined access policies Access Policy Built T2 Healthcare delegation Protected Delegation Built T2 Managed agent identity Agent Control Panel Built T2 Duress-resistant delegation Duress Safeguard Built T2 Containing a compromised agent Contain Agent Built T3 Review before commit checkpoint Draft Review Built T3 Confirmation receipt showing what happened Action Receipt Built T3 Audit trail with role-based views Audit Trail Built T3 Reasons for decision Reasons for Decision Built T3 Recourse and dispute resolution Dispute & Recourse Built T3 Showing liability at the point of action Liability Statement Built T3 Reversibility and undo Reversibility Preview Built T3 Circuit breaker for agent actions Circuit Breaker Built T4 Showing confidence and uncertainty Eligibility Check Built T4 Trust marks and certification Trust Mark Built T4 Graduated delegation Delegation Ladder Built T4 Transparency by default Transparency Moments Built T4 Communicating the level of automation Automation Label Built T4 Signaling human-in-the-loop oversight Oversight Signal Built T4 Earned trust and reversible delegation Earned-Trust Control Built T5 Rate limits per verified person Rate-Limit Meter Built T5 Structured intake that resists volume-padding Structured Intake Built T5 Proof of personhood without CAPTCHAs Personhood Token Built T5 Participatory budgeting as an alternative to rationing Participatory Budget Built T5 E-petition threshold design Petition Threshold Built T5 Sludge audits and how they backfire Sludge Audit Built T5 Appeal volume and overturn rate as early warning Appeals Dashboard Built T6 Clustering and deduplication for high-volume submissions Cluster Console Built T6 Provenance and attribution for mass submissions Submission Record Built T6 Opinion mapping by consensus and clustering Live Opinion Map Built T6 Weighting distinct voices Results Dashboard Built T6 Submitter nudges on template matches Template Nudge Built T6 Alternative weighting mechanisms Credit Voting Built T6 Processing submissions that resist prompt injection Review Pipeline Built T6 Verified agent admission Agent Admission Built T7 Agency-provided agents (a public option) Service Entry Built T7 Preserving the person's own voice Voice Fidelity Built T7 Non-agent fallback channels, kept first-class Channel Choice Built T7 Accessibility baseline: WCAG and beyond Accessibility Controls Built T7 Plain language and multilingual intake Language Intake Built T7 Co-design with affected communities Community Oversight Built T7 Whose interests the platform serves Interest Disclosure Built T7 Government-provided agents for high-stakes interactions Tiered Advocacy Built T7 Digital divide and exclusion risk Exclusion Assessment Built T8 Certification marks and trust registries Trust Registry Built T8 Nutrition labels for AI tools and datasets Nutrition Label Built T8 Software bills of materials for user-facing tools Supply-Chain Status Built T8 Risk-proportionate review of civic tools Submission Review Built T8 Lifecycle certification for tools that advise users Risk Classification Built T8 From voluntary framework to certifiable standard Framework Assessment Built T8 Grounding and source attribution in AI outputs Grounded Answer Built T8 Lightweight certification for the long tail Certification Wizard Built T9 Jurisdiction disclosure when data is processed Jurisdiction Badge Built T9 Data-sovereignty tiers for sensitive interactions Sovereignty Tier Built T9 AI system transparency registers Transparency Register Built T9 Cross-border data transfer as a design obligation Transfer Basis Built T9 Sector-specific data residency Residency Assurance Built T9 Sovereign AI model selection and disclosure Model Provenance Built T9 Concentration-risk and supply-chain disclosure Sovereignty Stack Built T9 Context-triggered disclosure Sovereignty Checkpoint Built