Thesis Frame
"Trust" appears in nearly every design leadership posting across your target list. OpenAI uses it. Maven Clinic uses it. Gusto uses it. Atlassian uses it. They mean four different things.
It fragments into four problems: delegation trust (AI-native), consequence trust (growth-stage regulated), coherence trust (enterprise platform), compliance trust (healthcare/vertical SaaS). Each tier hires for its version and screens against the others.
"I design trustworthy systems" sounds senior until the follow-up: trust in what failure mode? You either have a concrete answer grounded in that company's specific problem, or you're exposed. Leading with delegation trust vocabulary at a company solving consequence trust reads as a precise-sounding answer to the wrong question. The vocabulary lands. The failure mode is off by a tier.
This is the most common positioning failure I see from design leaders with AI experience at Director+. The rest of this section gives you the correction: one playbook per trust type, a cross-tier evaluation guide, and company-level decoder cards.
Same Portfolio, Four Different Candidates
Your Alibaba case enters four rooms and becomes four different candidates. An enterprise panel sees coherence at scale. An AI-native evaluator scans for structural patterns about human-agent boundaries. A growth-stage platform sees the logo and worries you need a large org to operate. Before any of them open your portfolio, a search partner has already compressed you into a phrase — and that phrase follows your name through every evaluation stage. This piece maps how each tier screens for trust competence, where the real gate falls, and which tag you need installed before you walk in.
The AI-Native Mandate Test
Pull up any two AI-native job postings with "design" in the title. Read the verbs. One says shape, define, evaluate. The other says craft, build, ship. Same title. Structurally different mandates. One puts you in the room where model behavior gets decided. The other hands you the output and asks you to make it usable. This playbook reads eight companies—OpenAI, Anthropic, Suno, Cartesia, DeepMind, Tools for Humanity, Plaud, Heidi Health—posting by posting, so you can detect which mandate you're actually being evaluated for before you spend outreach energy on the wrong one.
The Exception-State Playbook
Every design portfolio at the Director+ level showcases success states. Clean dashboards, smooth onboarding, polished outcomes. Headway, Gusto, and Ramp are scanning for something else. Their evaluation rooms are looking for evidence you've designed for the moment the system is wrong — wrong billing, failed verification, late payroll — in contexts where "wrong" has a dollar amount and a paper trail. This playbook maps the tier's shared evaluation DNA from current postings and shows you which evidence to lead with, which framings will get you killed, and how to read whether the company deserves you back.
Enterprise Platform Tier Playbook — Consistency as the Trust Problem
Atlassian is rebuilding its design system as AI-readable infrastructure. Microsoft frames the Head of Design role as editor-in-chief of a coherence problem spanning Copilot, Office, and developer tools simultaneously. Intuit's live Director posting says it directly: "from strong individual features toward a cohesive, high-quality platform." Enterprise design leadership is being redefined around consistency as a trust problem, and AI modernization is compressing the timeline. Four postings map to your background right now. This playbook covers positioning, the mandate trap that kills external candidates, and the red flags the companies show you.