The physical workplace has not fully caught up to the AI-centered office. For many organizations, the conversation still begins with software adoption: which tools to approve, how to govern data, how to train employees, and where automation can reduce low-value work. Those questions matter. Yet they leave out a practical reality: when the way people work changes, the places that support work need to change as well.

Steelcase’s recent article, Creating People-Centered AI Spaces, makes a timely case for rethinking the office around new AI-enabled behaviors. The piece cites research showing that 78% of U.S. leaders believe AI will lead to office redesign within the next few years, while 35% of AI users say AI is already changing how they use the physical workplace.

The more useful question is not whether AI belongs in the workplace. It is where AI is already changing work patterns, and what the office must do differently as a result.

For many employees, AI is becoming part of the daily work sequence. A person may use AI privately to draft, summarize, compare options, or pressure-test an idea before bringing that work to colleagues. Team time may then shift toward review, critique, synthesis, and decision-making. Afterward, AI may help translate the discussion into notes, action items, or follow-up communication.

That sequence changes the role of the office. The workplace is no longer a destination for individual focus or scheduled meetings. It becomes the connective tissue between individual AI-assisted work and human judgment.

This is where design decisions become business decisions.

The AI-enabled workplace is a connected system of people, place, furniture, technology, and change management. Value comes from understanding how work actually moves during the day, then designing settings that support those transitions with less friction.

Focus spaces are a good example: traditional focus rooms were often designed around quiet and privacy, which are still necessary. But AI-assisted focus work adds new requirements. People may need to speak naturally to an AI tool, compare multiple sources on screen, review generated content, or move quickly from private exploration to a conversation with a colleague. That means privacy, acoustics, display support, power access, lighting, and adjacency all need to be considered together.

A small enclosed room with poor ventilation, one undersized monitor, and weak sound control may technically count as a focus space. It will not support this new pattern very well.

The same is true for collaboration. AI may help teams enter meetings with more information, but more information does not automatically create better decisions. In some cases, it can create the opposite problem: polished material that appears complete before the thinking has been tested. Teams need places where they can examine AI-generated work, compare it against real context, challenge assumptions, and decide what should move forward.

That requires project rooms and collaboration settings that are more flexible than the traditional conference room. Teams may need large shared displays, writable surfaces, mobile furniture, reliable content sharing, and room layouts that support both in-person and remote participation. They also need conditions that reduce cognitive strain: clear sightlines, consistent audio, appropriate lighting, and intuitive room controls.

This is where workplace technology matters most. Cameras, microphones, speakers, displays, cabling, acoustics, and control systems all influence whether a room supports the work or interrupts it. As AI-enabled capture, transcription, translation, and summarization become more common, the quality of the room environment will affect the quality of the digital record. If a remote participant cannot hear side comments, if a camera misses the person speaking, or if the room layout privileges people at the table over people on screen, the technology may reinforce the very inequities it was meant to reduce.

The workplace also needs to account for the human cost of more screen-based work. Steelcase notes that heavier AI use can correlate with higher stress and burnout. That should not be treated as a secondary issue. If employees spend more of the day in high-intensity digital interaction, then social and restorative spaces become part of the workplace performance strategy.

These spaces do not need to be elaborate, but they do need to be intentional. A well-designed cafe, lounge, library setting, or informal neighborhood hub can give people a place to reconnect, step away from constant digital input, and build the trust that makes collaboration more effective later. Human connection is not a soft benefit in an AI-enabled office. It is part of the operating model.

For leaders, the opportunity is to move beyond the idea of an “AI room” or an “AI office.” AI will continue to change too quickly for static solutions. A better approach is to create a workplace ecosystem that can adapt as behaviors change.

That starts with observing how work is actually happening. Where are people using AI now? Which tasks have moved from group settings to individual exploration? Which meetings now require more review and validation? Where are employees struggling with privacy, noise, display access, remote equity, or room technology? Where are they improvising because the workplace does not support the flow of work?

The answers will vary by organization, industry, and team. A professional services firm may need more project rooms for fast-moving client work. A technology company may need a stronger mix of individual focus spaces and team synthesis zones. A healthcare, education, or public sector workplace may need to pay closer attention to privacy, governance, and equitable participation.

The principle is consistent: design for behavior before designing for tools. AI may accelerate work, expand access to information, and change how teams create and evaluate ideas. But people still provide the judgment, trust, context, and accountability that make work meaningful.

The future workplace should not be designed around AI as an object of attention. It should be designed around the new patterns of work AI is creating, and around the people who still have to decide what is true, useful, responsible, and worth doing.