AI Tools for Real Estate Brokerages: What to Pay For

By Emily Terrell — Top Coach and Speaker at Tom Ferry International. Speaker for NAHREP, eXp Con, and brokerages nationwide.

The AI tools for real estate brokerages worth paying for are one business-tier AI workspace, such as ChatGPT Business or Claude Team, plus the AI features already inside your CRM. Business tiers keep agent and client data out of model training by default. This guide covers who gets a seat, what to stop reimbursing, and the cost math.

Key Takeaways

• Buy one governed AI workspace for the brokerage, not a stack of reimbursed personal subscriptions.
• The reason to pay for a business tier is data handling and admin control, not a smarter model.
• Turn on the AI you already own in your CRM before you buy anything new.
• Seat the people who touch client data first, then expand based on usage, not attendance.
• Software without an internal owner and a review gate is a line item, not an adoption plan.

What is a brokerage AI stack?

A brokerage AI stack is the set of AI tools a firm pays for, approves, and supports on behalf of its agents and staff, as opposed to the tools agents buy on their own. It has three layers: a general-purpose AI workspace, the AI features inside the CRM and productivity suite, and the policy and training that govern both. Whatever the brokerage provides becomes the default most agents actually use.

Why this matters for brokers and real estate agents

Your agents already believe you’ve handled this. According to NAR’s 2025 Technology Survey (September 2025), two out of three agents either agree (38%) or strongly agree (29%) that their brokerage provides all the tech tools they need. That same survey found 46% of agents said AI had no noticeable impact on their business, and 24% spent over $500 a month on tech tools.

Read those together. Most agents trust the brokerage stack, nearly half get nothing measurable from AI, and a quarter are paying heavily out of pocket anyway. Here’s the thing nobody wants to tell you: in my coaching, the agents getting nothing from AI are usually working alone, on personal accounts, with no shared prompts and no review gate. That’s a brokerage design problem showing up as an agent performance problem.

The cost pressure is real on both sides of the desk. According to NAR’s 2026 Member Profile (June 2026), median business expenses for Realtors rose to $9,530 in 2025, up from $8,010 the year before, and 86% of Realtors are independent contractors at their firms. You can’t mandate a contractor’s tools. You can make the approved option the easiest and cheapest one available.

NAR’s own broker guidance points the same direction. Its March 2026 article on brokerage AI policy tells brokers to create a list of approved tools, especially where client or transaction data is involved. An approved list only works if the brokerage pays for something worth approving.

“A brokerage that reimburses twenty personal AI subscriptions doesn’t have an AI strategy. It has twenty AI strategies and no idea where the client data went.”
— Emily Terrell, Tom Ferry Coach

The five-question brokerage AI buying framework

Run these in order. Each answer narrows the next decision, and most brokerages find they need to buy less than they expected.

What does your stack already include?

Audit what you already pay for before you add a vendor. Most brokerages own more AI than they’ve switched on.

Start with the CRM, because it’s the layer attached to revenue. Follow Up Boss, for example, lists AI features as included at no additional charge, with smart summaries, smart messages, suggested tasks, and predictive lead prioritization on its Pro and Platform plans. On the entry Grow plan, some of those features rely on data from Calling, a paid add-on. Then check your productivity suite. If the firm runs on Google Workspace or Microsoft 365, find out what AI your current plan already includes before anyone signs a second contract.

Which tool is allowed to touch client data?

Only a business-tier tool with documented data terms, and that’s the real reason to pay at all.

OpenAI’s pricing page states that ChatGPT Business does not train on your business data by default. Anthropic’s pricing page lists no model training by default on Claude Team, while its consumer plans, including Pro, are listed as opt-out. That difference is the whole argument. A consumer subscription, even a paid one, puts the data decision in each agent’s settings menu. A business workspace puts it in yours.

The account tier is only one layer. Redaction rules and approved use cases still matter, and I’ve broken down what’s actually safe to put into ChatGPT separately. This is general information, not legal advice. Confirm your obligations with your broker of record and an attorney licensed in your state.

Which workspace should you standardize on?

Pick one, and pick the one your agents already open. The two leading business workspaces cost the same, so price won’t decide this for you.

ChatGPT Business is $20 per standard seat per month billed annually, or $25 billed monthly, for teams of 2 to 200. Claude Team is $20 per standard seat per month billed annually, or $25 billed monthly, for teams of 2 to 150. Both include centralized billing, admin controls, and single sign-on.

Adoption is the tiebreaker. NAR’s 2025 Technology Survey found ChatGPT was the most-used AI tool among agents at 58%, ahead of Gemini at 20% and Copilot at 15%. If your agents already live in ChatGPT, that habit is worth more than any feature comparison. Choose Claude when your use case is long documents and consistent brand voice across a team’s content, and you’re willing to train for the switch.

What you should not do is buy both “so agents can choose.” Two workspaces means two prompt libraries, two sets of settings to audit, and half the shared knowledge in each.

Who gets a seat first?

The people who touch client data every day, then the agents already using AI weekly. Not the whole roster on day one.

Your transaction coordinators, admins, marketing staff, and ISAs handle the most sensitive files and produce the most repeatable work. Seat them first. Then add producing agents who already use AI. NAR’s 2025 survey found 20% of agents use AI tools daily and 22% weekly. Those are your early adopters, and they’ll build the shared prompts everyone else inherits.

Both workspaces include usage analytics. Review them quarterly and move unused seats to agents asking for one. A seat that hasn’t been opened in 30 days goes back in the pool.

The math stays small. Fifteen standard seats on annual billing run about $300 a month on either platform. That’s less than what one agent in the survey’s top spending bracket pays for tech tools alone.

What should you stop reimbursing?

Personal AI subscriptions and any tool that duplicates what the brokerage now provides.

Once a governed workspace exists, reimbursing consumer plans undercuts it. You’re paying for the tools your policy says not to use with client data. The same goes for standalone AI writing tools that draft follow-up your CRM already drafts. For the agent-side version of this audit, point your people to how much agents should budget for AI tools. It gives them the same discipline at the individual level.

Video and design tools are different. Agents choose those based on how they market, and as independent contractors, that’s their call. Approve categories, not every app.

How I use this in my own business

My own team runs on one shared AI workspace, not a collection of personal logins. My transaction coordinator and my VA have seats because they touch client files every day. The prompt library lives in that workspace: listing intake, seller updates, comp narratives, and the redaction template that strips identifiers before anything gets pasted. A new team member starts with the system on day one instead of rebuilding it from a group text.

That’s a big part of how I close 70+ transactions a year in San Antonio on roughly five hours a week of active management. The tool isn’t special. The structure around it is: one workspace, named owners, and a review gate before anything client-facing goes out.

When I work with brokerage leaders, the first exercise is the same one I ran on my own business. List every AI charge the firm and its agents are paying for, name the task each one removes, and cut everything that can’t answer. Most firms find the governed workspace costs less than the sprawl it replaces.

Common mistakes

Buying seats for the whole roster on day one. Half of them won’t be opened in the first month. Start with staff and weekly users, then expand based on the usage report.

Running two workspaces to keep agents happy. You’ll split your prompt library, double your settings audit, and get half the shared knowledge in each.

Paying for a business tier and leaving the CRM AI switched off. The CRM is where follow-up actually happens. Turn on what you already own first.

Approving a tool with no review gate. NAR’s broker guidance is direct that AI-generated content shouldn’t go straight from prompt to publication. Your approved tool needs an approved process. Build it into your own AI policy, not just the brokerage’s.

Treating the purchase as the rollout. Seats don’t create adoption. A named internal owner and a 90-day cadence do, which is why most brokerage AI training doesn’t stick without one.

Frequently Asked Questions

What AI tools should a real estate brokerage pay for?

One business-tier AI workspace, such as ChatGPT Business or Claude Team, plus the AI features already included in your CRM and productivity suite. Business tiers exclude your data from model training by default and give you admin controls. Everything else, including video and design tools, can stay an agent-level decision, since most agents are independent contractors who choose how they market.

Should a brokerage reimburse ChatGPT Plus for its agents?

Usually not. Plus is a consumer subscription, so the data and training settings live in each agent’s account rather than under brokerage control. For the same $20 to $25 per seat, ChatGPT Business adds a shared workspace, centralized billing, admin controls, and no training on business data by default. Reimbursing personal plans undercuts any approved-tool policy you write.

Is ChatGPT Business or Claude Team better for a brokerage?

They’re priced the same, so choose on workflow fit. ChatGPT has the adoption lead, with NAR’s 2025 Technology Survey finding it was the most-used AI tool among agents at 58%. Claude suits teams producing long documents and consistent brand voice across a body of content. Pick one and commit, because running both splits your shared prompt library.

How much does a brokerage AI workspace cost?

ChatGPT Business and Claude Team both list standard seats at $20 per user per month billed annually, or $25 billed monthly, with a two-seat minimum. Fifteen seats on annual billing run about $300 a month. Check each vendor’s pricing page before you buy, because AI plans and prices have changed repeatedly over the past year.

Can a brokerage require agents to use a specific AI tool?

Not in most cases. According to NAR’s 2026 Member Profile, 86% of Realtors are independent contractors at their firms, so a brokerage can’t mandate their tech stack. What a brokerage can do is set rules for client and transaction data, approve specific tools for that data, and make the approved workspace the easiest option.

Does a business AI plan make client data safe?

It removes one risk, not all of them. Business tiers from OpenAI and Anthropic exclude your data from model training by default, but agents still need redaction rules, approved use cases, and human review before anything client-facing goes out. This is general information, not legal advice. Confirm your obligations with your broker of record and a licensed attorney.

Who should get an AI seat first at a brokerage?

Staff who handle client files every day, such as transaction coordinators, admins, marketing coordinators, and ISAs. Next, add agents already using AI weekly. NAR’s 2025 Technology Survey found 20% of agents use AI tools daily and 22% weekly. Review usage quarterly and reassign any seat that hasn’t been opened in 30 days.

Bring this to your team or event

Emily Terrell speaks at brokerage events, real estate conferences, and team trainings on AI, systems, and social media — the exact playbook in this post, delivered live to your audience. As a Top Coach and Speaker at Tom Ferry International and an active agent closing 70+ transactions a year, Emily speaks from the stage about what’s working right now, not theory. Recent stages include NAHREP and eXp Con.

Book Emily to speak at your next event:
Email: eterrell@yourcoach.com
Phone: (210) 400-9191
Web: coachemilyterrell.com

For real estate agents who want to implement this: Get the weekly real estate prompt library at weeklyrealestateprompts.com or follow @coachemilyterrell on Instagram for daily systems and AI breakdowns.