2025 in Review: Design Gives AI Meaning

Source: Hsiao-Ching Chou; Salesforce Design · 2025-12-17

Design sets the foundation for trustworthy AI.

Designers help establish the semantic foundation that teaches AI what things mean and how humans work.

Design is a multiplier that transforms AI potential into practical business value.

In 2025, while the headlines shouted AI breakthroughs, designers were solving a deeper question: How do you architect AI’s ability to understand the world and the ways that humans work – especially in the context of Salesforce?

While the company focused on Agentforce, we zoomed in on the underpinnings of how AI can support humans. The decisions that determine ease of use, and thus adoptability, are rooted in design – though it’s not always apparent.

The responsibilities of design have expanded in the age AI. We learned from Kat Holmes, our Chief Experience Officer at Salesforce, that we’re “building expertise in agent experience design, where AI agents are treated as a new kind of user.”

In her blog, “Welcome to a new Era of Experience with AI Agents,” she discussed the design of agents and for agents to ensure AI prioritizes the goals of human users who are becoming agent orchestrators. Because AI systems are dynamic, there’s no need to have a single user journey or click path. Flexibility is key as people “discover what makes AI agents feel like an embedded and invaluable part of how” work gets done.

To borrow the words of a colleague who’s a principal AI architect, design used to be about creating screens. Now, it’s about deciding what happens before the screen appears.

Diving a bit deeper, Principal Product Designer Hervé Mischler drove the point that traditional UX design doesn’t have frameworks for coding assistants and multi-agent orchestration. Thus, there’s a UX paradigm shift that requires designers to embrace agentic experience design, system building, and creating “experiences that harness coordinated AI agents across platforms while maintaining human agency and understanding.”

Design Director JoEllen Kames, on the Agentforce platform team, explained the importance of experience architecture and runtime experiences. To make dynamic systems possible, product designers have to architect the “invisible” scaffolding that enables AI systems to respond in real time and “create experiences that surface the right things at the right time and in the right places.”

The proliferation of generative AI caused many practitioners to take stock and upskill. Learning designer and writer Nic Dimond shared a series of blogs reassuring designers that their UX expertise translates well, and that they can create a learning path to understand key agentic concepts.

Of course, with “vibe coding” sweeping across the industry this summer, many designers and developers began experimenting with coding assistants like Cursor and Windsurf to quickly generate prototypes. (That early experimentation was a helpful prelude to the fall launch of Agentforce Vibes for enterprises.) Vibe coding blurred the line between designers and developers, expanding their mutual capabilties.

Necessarily, Kat Holmes implored CEOs to invest in their teams and focus on four critical design skills to thrive with AI: experience architecture, designing agent collaboration, conversation design, and building context maps and ontologies.

The SLDS 2 launch in 2025 delivered the foundation for agentic design systems for Salesforce products built on the Lightning Platform. The decoupled architecture, modular components, agentic design patterns, and global styling hooks make it easier for users to build and customize dynamic experiences at scale. The most recent releases include dark mode (beta) and Salesforce DX MCP.

It’s systems thinking, explained former Salesforce designer Min Chang, that helps him understand the concepts that shape how a product works. “Instead of designing for isolated screens, I’m now designing how a system moves between states, like logged in or logged out, light mode or dark mode.”

Speaking of systems thinking, it’s a core design skill. Lead Researcher Holly Prouty unpacked the growth opportunities for designers in describing how the Jobs to be Done for UX designers have changed. “There’s an increasing desire and capacity to spend energy ensuring the right things are being built. ...So, designers are spending more time connecting the dots across products, teams, and workstreams.”

Understanding human communication patterns and expectations is essential to building AI that responds to the way humans work and collaborate. Conversation design is rich with insights that can shape AI behavior. We explored “what is conversation design” and conversational UI, how to repair broken conversations between humans and AI, and the impact conversation design makes to the quality of call with an AI agent.

Understanding human expectations of AI affects its design. Our researchers Michelle Tabart and Mar Drouhard have led studies on how people perceive the value of agents and whether Agentforce improves productivity for your sales team. It turns out, just because your AI agent works doesn’t mean your teams will find it useful.

In a related topic, Principal Design Strategists Andrea Small and Kate Piper, with Experience Strategy Architect Bethany Pickard, dug into the ways that humans collaborate at work, which informs their expectations of AI. Currently, most tools tend to support independent work. The race is on to design “multiplayer” experiences that enable teams to collaborate more easily with one another and AI agents.

AI’s powerful capabilities can easily go sideways if the machine doesn’t understand the language and context of your business. This is where ontology – a system for organizing the data about the data – is crucial to experience design. As Principal Ontologist Madonnalisa Chan shared, without ontologies, your experience using any number of digital platforms, such as Netflix or Spotify, would be chaotic.