The three facets of AI for design

31 Aug

AI is moving past hype and into infrastructure. Like mobile and cloud before it, AI is reshaping how design delivers impact across workflows, products, and platforms. For leaders, there’s an immediate opportunity to embed AI into the systems, teams, and practices that will define the next decade of design.

AI isn’t one thing. It shows up in three overlapping spaces leaders must account for:

AI augmentation for workflows and process
AI products and integrations
AI as a platform

Here’s what each means, and what you can do now to build momentum.

1. AI augmentation for workflows and process

This is where most design teams start. AI speeds up how work gets done:

Research analysis becomes faster when AI can cluster interviews or parse sentiment.
Copy and content variations can be generated instantly, giving designers more time to refine tone instead of starting from scratch.
Visual explorations or wireframe options can be AI-assisted, freeing up headspace for strategic framing.

The catch: without guardrails, this quickly turns into an AI dumping ground without clear direction or goals. Leaders have to shift teams from using AI to produce noise into using it to amplify judgment, context, and expertise.

What to do:

Create an AI augmentation playbook - what tools are approved, where they fit, how to use them, what guardrails exist
As part of your playbook, develop a prompt library that designers can use off the shelf.
Assign an AI workflow lead inside UX Ops to pilot, document, and scale best practices.
Run quarterly audits of how AI is being used - are designers getting leverage, or just spinning out noise? Where can improvements be made? How has technology changed and how should you update accordingly?

This guide is a very helpful starting point of all the things you may need to consider when bringing AI into your design workflows.

A brief call-out - there’s a whole subset of work around AI augmentation for design manager workflows, particularly around growth. I talk about that in this post.

2. AI products and integrations

This is the outward-facing layer: designing AI-powered features that customers use. It’s where UX teams partner with product and engineering to:

Decide how and when AI should show up in workflows.
Handle transparency, explainability, and trust — so users know why a recommendation or summary appeared.
Build patterns that keep AI outputs useful but human-centered.

Here, designers can’t just be interface decorators. They need to engage with system behaviors, ethical implications, and data flows. Good design makes AI usable; bad design makes it creepy or useless.

What to do:

Form an AI experience working group (design, PM, eng, data science, legal/ethics, and any other relevant stakeholders) to set principles for AI in product.
Build a design pattern library for AI interactions (explainability modules, feedback loops, opt-in/opt-out).
Train senior designers to map user journeys with AI as an actor in the system, not just as a UI element.

3. AI as a platform

This is the infrastructure layer — the one most designers forget exists. Companies don’t just ship features; they need an AI Platform team to handle:

Models, taxonomy, and data pipelines (what powers the features)
Governance and compliance (guardrails, auditability, regulatory risk)
Delivery systems (APIs, SDKs, integration points across the product ecosystem)

For design, this means two things:

Designers need a seat at the table to influence how AI platform capabilities are exposed. Otherwise, teams could end up with dead ends no one can design around.
Designers themselves need literacy in how platforms work - not to become engineers, but to understand the system architecture and data models that underpin user experiences.

What to do:

Advocate for a dedicated AI Platform team and secure design representation within it.
Stand up an AI architecture working group that includes designers, so UI decisions are linked to data/model decisions.
Establish AI design standards and principles. Extend your existing design principles to cover things like transparency, explainability, trust, and ethical use of AI, and create practical guidelines for when and how those principles apply in product.
Build a designer’s guide to AI platform literacy (APIs, SDKs, pipelines) to raise baseline fluency across the team.

My approach across all facets

AI belongs in the operational backbone of design. Not as a novelty or as a hackathon project, but as part of the system that scales. That means:

Positioning AI Ops alongside Design Ops and Research Ops. If you already have teams managing process, tooling, and insights, AI should live there too. It needs the same rigor and ownership.
Creating resilient systems and guardrails. Handing designers a bag of shiny AI tools and hoping for the best won’t cut it here. As we continue shifting designers even more away Figma pixel-pushing to broader problem-framing and strategic design, that requires well-defined workflows, governance, and quality standards.
Upskilling for impact. .Teams need to develop expertise in strategy, system architecture, data literacy, and timeless design principles. The future designer is less “make this button blue” and more “frame the problem so AI can help solve it responsibly.”

Why this matters

AI is no longer an add-on; it’s infrastructure. The leaders who recognize that, and design the systems around it, will shape not just better products but the next era of design itself.

platform designAIproduct design