Release of ChatGPT Enterprise – Implications for CRE

If you’ve been reading A.CRE lately, you know that we see huge potential in generative AI (e.g. ChatGPT for language, Midjourney for images, etc). We’ve committed an entire section of our site to AI in commercial real estate, created a series exploring use cases for AI in CRE, and even keep a running list of AI tools we think will have an impact in real estate. Or put another way, we’ve been an advocate for safe adoption of this nascent technology.

However, one big barrier to adoption has been privacy and security of proprietary data. While these models can be trained on a real estate companies’ vast dataset, doing so risks exposing that highly valuable data. Thus, most companies have been in a wait-and-see mode. I believe that changes today.

Today, OpenAI announced the release of ChatGPT Enterprise, with robust security, high-speed GPT-4 access, and advanced customization to elevate work efficiency. It’s a leap toward an AI work assistant for real estate that boosts creativity while safeguarding company data. In this post, we’ll share details of this new AI product and discuss the implications it has for commercial real estate.

What is ChatGPT Enterprise?

ChatGPT Enterprise is the latest offering from OpenAI, designed specifically for enterprise-level applications. It was officially released on August 28, 2023, with a focus on robust security, customization, and scalability. The platform offers enterprise-grade security features that are SOC 2 compliant, ensuring that all conversations are encrypted both in transit and at rest. Importantly, the data and conversations are not used to train OpenAI’s models, maintaining data privacy.

• Not sure what ChatGPT or large language models are? Read how ChatGPT is used to improved productivity in CRE.

The platform also allows unlimited access to the GPT-4 model, which performs up to two times faster compared to the standard version. Users can benefit from advanced data analysis capabilities, formerly known as Code Interpreter. The system supports longer context windows, allowing for more extended conversations and data inputs with a 32k token context.

For ease of deployment across large organizations, ChatGPT Enterprise includes features like an admin console for bulk member management, Single Sign-On (SSO), and domain verification. It also provides an analytics dashboard to gain insights into usage.

The platform is available for immediate adoption, with more features expected to be rolled out in the near future, including options for smaller teams and more specialized tools for various roles within an organization.

So, what are the issues and opportunities in commercial real estate of this tech?

The Security Quandary: A Closer Look

Data security has long been a stumbling block for CRE firms contemplating the adoption of generative AI technologies like ChatGPT. The hesitancy isn’t unwarranted. Commercial real estate is a data-rich industry, and that data is fiercely guarded. Imagine the consequences if a competitor gained access to proprietary information on asset valuations, tenant relationships, investment strategies.

With the public version of ChatGPT, there’s a persistent concern that user prompts could be used to train the model. This is more than just a theoretical risk; it’s a real issue that could inadvertently expose confidential data. Given that these models learn from the data they interact with, the chance of revealing sensitive information, albeit slim, has been enough to keep many CRE firms at arm’s length.

ChatGPT Enterprise changes the equation by promising that a business’ data stays its own. It doesn’t use prompt conversations for training, and with SOC 2 compliance, the security measures meet industry standards. This could be the assurance that CRE firms have been waiting for—a way to leverage AI’s capabilities without sacrificing the sanctity of their data.

Data Integration: Opportunity and Challenge

The prospect of ChatGPT Enterprise seamlessly syncing with a real estate firm’s data is compelling. We’re talking about faster decision-making, better collaboration, and advanced data analysis. But let’s not get ahead of ourselves. CRE data is often a messy mix of spreadsheets, databases, Argus files, CoStar exports, and ad-hoc records.

So, how well can an advanced tool like ChatGPT Enterprise handle this unstructured landscape? Features like “advanced data analysis” and “32k token context windows” are fantastic, but their effectiveness is conditional. Garbage in, garbage out, as the saying goes.

Here’s where opportunity knocks for CRE professionals: data organization. The ability to clean, structure, and standardize data will be a high-demand skill. The ability to implement AI into a CRE firm will be gold. If you’re looking to ride the AI wave in CRE, being a data maestro will put you ahead.

In short, ChatGPT Enterprise offers powerful tools, but their real-world application hinges on the quality of your data. Before embracing this tech, firms need to assess their data readiness. The tool’s potential is big, but it’s not a magic wand.

Winners and Losers in CRE from Gen AI

The advent of ChatGPT Enterprise sets the stage for a new dynamic in CRE: the formation of clear winners and losers in the race to adopt generative AI. On one side, you have firms sitting on a gold mine of proprietary data. Ironically, these are often the behemoths that are slowest to adapt to new technologies. Their hesitancy isn’t just a missed opportunity; it’s a liability.

On the flip side, you’ve got smaller, nimble firms that are quick to adopt new tech but may not have the same depth of data. The real winners in this scenario are the firms that can strike a balance—those that have both the data and the agility to adapt. These firms will leverage ChatGPT Enterprise to its fullest, gaining insights at speed and scale that others can’t match.

The losers? Those who resist change, relying solely on intuition and traditional methods. In a world where data-driven decisions are becoming the norm, flying by the seat of your pants is a risky strategy. Similarly, firms that lack the proprietary data to inform AI models will find themselves increasingly sidelined.

The message is clear: adapt or get left behind. In the context of ChatGPT Enterprise, having a wealth of data is a potent advantage, but only if you’re agile enough to use it. The stakes are high, and the divide between the tech-savvy and the tech-averse is widening.