Beyond the Buzz: Where Is AI Actually Delivering Value in CRE?

At WorkSpaces Napa, workplace and real estate leaders gathered for a closed-door discussion on one of the industry's most talked-about topics: AI.

These weren’t conference keynotes or polished soundbites. Just an honest, off-the-record conversation about what’s actually working, what’s falling flat, and where the potential really is.

In this deep dive session, the group unpacked the real friction points of enterprise AI adoption, from governance headaches and platform fatigue to unexpected wins with internal agents and content workflows.

Here’s a look at the insights, challenges, and use cases that came out of the room:

Policy vs. Productivity

The friction between innovation and governance is real.

Most enterprises have locked down AI usage with strict internal policies. That’s great for risk mitigation, but it’s also slowing teams down.

• One leader shared that their org mandates an internal enterprise AI, even when tools like Copilot or Gemini produce better results.
• Many teams are splitting usage: public tools for broad market research, internal tools for anything involving company data. It’s not ideal, but it works for now.
• Policies vary wildly. Some companies restrict teams to one tool, while others allow multiple, but lack clear guidance or training.

Across the board, leaders feel the pressure to adopt AI faster than internal policies can keep up.

Internal Agents & The Metadata Problem

Some companies are rolling out internal agents to help with tasks like booking rooms, managing budgets, and answering HR or finance questions. But the success of those agents depends on structured, accessible metadata.

Without it? Agents break down quickly and become more noise than help.

• Teams are seeing "agent sprawl", multiple bots doing similar tasks with no clear ownership.
• A few clever workarounds came up. For example, one service provider builds agents outside the internal platform to avoid legacy system constraints, still compliant, but faster and more accurate.

It’s Not the Tool. It’s the Training.

Several teams shared that adoption isn’t about throwing tools at people; it’s about changing how they work. Some tactics that helped:

• Peer-led sessions focused on one tool or use case
• AI “lunch & learns” during slow seasons
  Clear rules for experimentation so teams feel safe testing ideas

Even with good tools, friction shows up:

• A tool nicknamed “Fixer” showed real potential (drafting emails, managing calendars), but too many early bugs pushed people to abandon it.
  Platform changes (like moving off Slack) killed momentum and created confusion.
  Bottom line: AI adoption only sticks when teams are coached on how to use it in context.

Real Use Cases That Are Actually Working

These weren’t theoretical. Leaders shared specific, tactical wins that are already saving time and improving output:

Contracts + RFPs

• AI is helping legal teams compare contracts (like IFOA and ConsensusDocs) and align language in a single session.
• RFPs are now being run through AI to reduce bias and benchmark more consistently.

Content + Training

• Blunt feedback is rewritten into coaching language using AI.
• Dense onboarding docs are converted into short videos or audio formats with tools like Notebook LM.
• Internal explainer videos dropped in cost from $3,000 to $500 with AI-assisted production.
• Strategy docs are being rewritten in plain English for execs and adapted for multiple formats.

Workplace Asset Data

• Teams are estimating furniture or equipment specs from photos, avoiding site visits and speeding up planning.

Better Visualization = Faster Decisions

AI is helping cross-functional teams make decisions faster by embedding real-time insights into tools like Tableau.

• Teams are running “what if” models across HR, real estate, and finance, then jumping into ChatGPT to explore edge cases and build recommendations.
• Some orgs are hiring engineers to improve metadata structures and reduce duplicated agents across departments.

One team shared an unexpected win:
AI helped uncover why employees were spending time in their cars outside the office. By layering in facilities reports, the system revealed broken breakroom equipment as the cause — a simple fix that improved morale and space usage.

What’s Still Getting in the Way

Despite early wins, there are plenty of roadblocks:

• Tool abandonment is high, people revert to manual work when results aren’t consistent
• Poor data quality, especially in procurement, limits AI’s value
• Teams lack benchmarks for accuracy (when is “good enough” good enough?)
• HR, finance, and CRE data are still siloed, blocking bigger-picture insights
• Onboarding and training are inconsistent across roles

The group agreed: AI’s biggest challenge isn’t the technology. It’s organizational alignment.

Final Word

AI is picking up real momentum, but only for the teams willing to fix broken processes first. If your data’s messy, your systems disconnected, or your tools underused, AI will only make those gaps more visible.

Want to be in the room for conversations like this?

Request an invite to WorkSpaces →