Compliance-First Design Mindset

Design Governance
for Regulated Environments

Regulatory strategy is the domain of Legal Counsel. My job is to translate legal opinions into deterministic UI logic — components that carry the regulation's intent into every render, every event log, every audit trail. Compliance isn't a project phase; it's an architectural choice made at wireframe stage.

5+ Years · ASIC-regulated environment
40+ Jurisdictions · component-mapped
8 Regulatory updates absorbed · 3-5 day cycle
0 Design-related findings · 2 yr
ASIC AU FCA UK FINRA US ESMA EU MAS SG SEC US

Role Clarity

Where design meets Legal — three boundaries I respect

"Compliance-first design" doesn't mean designers play lawyer. It means designers build systems that carry Legal's intent into every render. The boundaries below are what keep that division of labour honest.

LEGAL OPINION TRANSLATE UI COMPONENT if (jurisdiction === 'AU') {  render(ASIC_warning)  require(scrollDepth) } EXECUTION OVER INTERPRETATION

Boundary 01

Execution over interpretation

My role isn't to read regulations — Legal does that. My role is to turn their opinion into deterministic UI logic: components that fire the right disclosure for the right jurisdiction every time, no human-in-the-loop required.

TECHNICAL BRIDGE
RiskDisclosure component · v3.2 ASIC RG 268 FCA COBS SEC 17a-4 MiFID Art 24 DATA-SWAPPABLE · ONE SHELL · 4 VARIANTS

Boundary 02

Knowledge in components, not memory

I don't memorise every clause. I build constraint-mapped component libraries — one shell, multiple jurisdictional variants as data swaps inside a unified state machine. Regulation changes? Update the data, not the code.

3-5 DAY UPDATES
JOINT FIGMA · DAY 1 wireframe prototype production DESIGN LEGAL → both in the file REVIEW DURING WIREFRAMING

Boundary 03

Cross-functional partnership · early

In regulated environments, design success is measured by risk mitigation, not aesthetics. I trigger Legal review during wireframing — not pixel-perfect handoff. 3 hours of friction at the start prevents 3 weeks of rework at the end.

LEGAL-FIRST

Methodology

Four questions I ask before a single pixel is drawn

Reactive correction is expensive. Proactive deterministic UI logic is architectural. Each question below shifts compliance work from "review after the fact" to "constraint baked in at the component level."

REQUESTED · 12 FIELDS FILTER NEEDED · 4 FIELDS if not needed, don't render

Question 01

Data minimisation · is this field critical?

Privacy-first architecture means defining PII boundaries before drawing pixels. If a field isn't needed for execution, it shouldn't exist in the UI state. Less data on the page means less data in the breach.

PII-CONSCIOUS
DARK PATTERN Pre-checked Marketing on Cookies all no audit trail POSITIVE ACTION Pre-checked Marketing on Cookies all logged + audit-ready

Question 02

Explicit consent · no auto-enabled traps

Components follow positive-action logic. Marketing opt-ins, cookie consent, risk-warning acknowledgments — all require explicit, logged user intent. Audit-readiness comes from the UI, not the privacy policy.

POSITIVE-ACTION
UNIFIED STATE MACHINE RiskDisclosure(jurisdiction, …) one component · data-swap variants AU UK US EU 40+ JURISDICTIONS · ONE CODEBASE

Question 03

Jurisdictional modularisation · ASIC ≠ FCA ≠ SEC

Not monolithic flows — constraint-mapped component libraries. ASIC, SEC, FCA variations are data swaps within a unified state machine. Regulatory deployment across 40+ countries: 3-5 days, not 3-4 weeks of rebuilds.

DATA SWAPS · NOT REBUILDS
EVENT LOG · USR_84371 14:32:18.247 disclosure_render v3.2 · ASIC 14:32:22.491 scroll_depth 100% · 4.2s dwell 14:32:25.118 consent_logged click · IP · UA 14:32:25.119 copy_hash a3f2…b91c 14:32:25.123 trade_executed precondition: ok DETERMINISTIC EVENT LOGS

Question 04

Evidence architecture · audit trails by design

Regulators require proof, not intent. UI generates deterministic event logs: disclosure render timestamps, scroll-depth verification, consent click + IP + UA, hash of the exact copy shown. Re-renderable 18 months later for SEC/ASIC inquiry.

SEC 17a-4 READY

Architecture

Compliance is a continuous state, not a project phase

In institutional finance, "audit-readiness" means a regulator can ask what a client saw 18 months ago and the system reconstructs it byte-for-byte. Two architectural patterns make that possible.

Order Ticket · BUY audit-id: ord_8a3f2b91 1,000 AAPL @ $187.42 hover_warning 4.2s scroll_to_bottom 100% click_acknowledge 14:32:25.118 FULL INTERACTION STATE · IMMUTABLE

Pattern 01

Deterministic interaction logging

Every component carrying compliance liability (order ladders, risk toggles, KYC fields) is mapped to a unique audit ID. We capture not just the final click but the full interaction state — hover dwell, scroll depth, time-to-acknowledge. Immutable, ready for regulator inquiry.

UNIQUE AUDIT ID
DISCLOSURE v3.2 · 2024-08-14 SHA-256: a3f2b91c8d…7f4e2c91 re-renderable EXACT COPY · 18 MONTHS LATER · 100% FIDELITY

Pattern 02

The "truth state" machine

A versioned disclosure system: the exact copy shown at execution time is cryptographically hashed and stored. If a regulator asks what a client saw 18 months ago, we re-render the exact UI state. Intent is ephemeral; evidence is architectural.

SHA-256 · VERSIONED

Resilience

Designing the violation out of the system

User training and manual oversight are the last line of defence. The first line is the UI itself — components that can't produce a violation, even if a user tries.

▲ ASIC RG 268 · DISCLOSURE CFDs are leveraged. You may lose more than your deposit. ○ rendered: yes ○ scrolled: 0% SUBMIT (disabled) precondition not met

Pattern 01

Conditional action locking

Order execution buttons stay disabled until the system verifies (1) the required risk disclosure has been rendered AND (2) the user has performed the mandatory scroll-to-bottom acknowledgment. The legal precondition is met before action is possible.

PRECONDITION GATE
NORMAL STATE BUY SELL Margin used: 35% ● Healthy 95% MARGIN WARNING CLOSE ONLY Margin used: 95% ▲ At-risk

Pattern 02

Automated market-state enforcement

If an asset enters a regulatory halt or an account hits a margin-warning threshold, the UI dynamically reconfigures. Market-order inputs are replaced with "Close Only" or risk-management interfaces — eliminating the cognitive load (and violation surface) during high-stress events.

DYNAMIC RECONFIG

Production Evidence

Two real cases — one regulated, one consumer

The methodology above isn't theoretical. Both examples below shipped, both went through Legal review, both became reusable templates.

Case 01 · ACY Securities

ASIC leverage restriction update · 2023

What happened: ASIC updated leverage limits for retail traders. Legal provided new disclosure requirements with a 14-day implementation deadline.

My role: I didn't interpret the regulation — Legal told me "this warning must be shown before every trade for Australian users." I designed the UI flow: modal disclosure → scroll-to-bottom acknowledgment → logged consent → trade execution. Shipped in 12 days.

Outcome: Legal confirmed design met ASIC requirements. The component became a reusable template for future regulatory updates — every subsequent ASIC change shipped in 3-5 days, not 14.

ASIC RG 268 SHIPPED 12 DAYS

Case 02 · PawsRoam

Age verification design · Japanese youth-protection

The question I asked: "If a 14-year-old wants to book a pet sitter, what could go wrong?"

Research: Japan's youth internet protection law and adjacent regulatory frameworks. Key takeaway: content filtering for under-18 users.

Design: Age gate at registration (birthdate verification). Under-18 accounts automatically hide late-night services, alcohol-related venues, unsupervised travel options. Parental dashboard for booking history.

Result: Compliance architecture supporting youth-protection and data-protection frameworks. Not because I'm a regulatory expert — because I asked the right questions early.

YOUTH PROTECTION PROACTIVE INQUIRY