Artificial intelligence has come a long way—but so have its blind spots.
According to a recent New York Times report, some of the most powerful AI tools in the world—ChatGPT, Google Gemini, DeepSeek—are hallucinating more than ever. And we’re not talking about small mistakes.
We’re talking about made-up company policies, fabricated facts, and confident answers that are flat-out wrong.
For those of us building companies and leveraging AI to grow, that should sound the alarm.
At BetterWho, we work with property management entrepreneurs using AI to scale—without sacrificing accuracy, credibility, or client trust. And this latest wave of news is your reminder: AI can help you move faster—but only when you stay in the driver’s seat.
Let’s break it down.
What Are AI Hallucinations?
AI hallucinations happen when systems like ChatGPT or Gemini generate information that’s inaccurate or completely fictional—but make it sound authoritative.
The NYT article we summarized here points to a real-world example:
A customer support bot from Cursor, a dev tool company, falsely told users they couldn’t use the product on more than one computer. Users canceled subscriptions. The company’s CEO had to step in on Reddit and say: “We have no such policy… This is an incorrect response from our AI support bot.”
Now imagine that in your world.
What happens when an AI chatbot misquotes your lease terms?
Or says your tenant service policy is different from what’s actually in your agreement?
Or generates an owner email with the wrong fee structure?
It’s a trust-breaker—and in property management, trust is everything.
Why Are AI Hallucinations Getting Worse?
The New York Times highlights a surprising trend: The newer and “smarter” the AI, the higher the error rate.
OpenAI’s o3 model hallucinates 33% of the time on benchmark tests.
Google and DeepSeek’s reasoning systems show similar, or even higher, error rates.
Some tasks, like summarizing news or handling complex reasoning, result in hallucinations up to 79% of the time.
Why is this happening?
Because today’s models are trained to “think” through problems step-by-step (aka reasoning), which creates more opportunities to hallucinate at every stage.
What This Means for Property Managers
Most property managers are underestimating the risk of bad AI.
They’re adding AI chatbots to their sites, using tools like ChatGPT to draft emails, or letting team members create documentation—without vetting the results.
That’s a fast track to misinformation, broken client trust, and real business damage.
When AI gets it wrong, your client doesn’t blame the tool—they blame you.
5 Ways to Use AI Safely in Your Property Management Business
At BetterWho, we teach clients and remote team members (RTMs) how to use AI for growth without losing control. Here’s how to make that happen:
1. Use AI for Drafting, Not Finalizing
AI is perfect for generating first drafts. But when it comes to client communications, leases, or fee explanations—always review and edit manually.
2. Use Prompt Templates with Guardrails
Instead of ad-libbing prompts, build standard templates that tell AI:
What documents to reference
What topics to avoid
How to phrase key policies
We help clients create AI-ready SOPs and checklists that reduce errors. We have also created an Ops Assistant called BettyWho that can create that for you.
3. Use AI in Low-Risk Areas
AI is great for:
Drafting maintenance reminders
Creating marketing content
Summarizing internal meetings
Avoid AI for:
Legal or financial messaging
Policy explanations
Owner fee discussions
4. Train Your Remote Team Members (RTMs)
Just giving your RTM access to ChatGPT isn’t enough. We train our RTMs to:
Ask for sources
Double-check lease or legal content
Use templates and structured review processes
Learn more about how our RTM Connect service helps clients scale smart.
5. Build AI Into Your Workflows
Don’t leave AI use to chance. Define where and how it’s used inside your business.
Example:
SOP: Responding to Owner Inquiries
Step 3: Use Prompt Template #2
Step 4: Send for internal approval before delivery
The BetterWho Approach to AI
We believe AI is the most powerful tool available to property managers today—but only when paired with leadership, training, and clear systems.
That’s why our coaching, consulting, and RTM placement services aren’t just about delegating work—they’re about building operational excellence that keeps you in control.
Your Quick AI Audit Checklist
- Do you know where AI is used in your company?
- Are your team members using prompt templates?
- Is AI output reviewed before going public?
- Are your RTMs trained to catch hallucinations?
- Do you have SOPs guiding AI usage?
Don’t Let AI Hurt Your Business
There’s no question AI can help you grow—but only when you use it wisely.
Don’t let hallucinations ruin your brand or cost you clients.
Let BetterWho help you build an AI-powered property management business that scales smart—with the systems and support to match.
Need help implementing a clean, smart AI strategy for your team?
Book a Free AI Audit call and let’s make AI your advantage—not your liability.
