How generative AI is transforming what’s possible in commercial real estate

• Acquisitions. When potential acquisitions are considered, GenAI can enable automation and intelligent analytics for finding, buying and operating assets. The due diligence effort becomes less onerous and more reliable, providing important insights that either support making the move or reveal concerns that prevent a transaction that shouldn’t be made. GenAI can also enhance portfolio planning, scenario planning, contracting and financial analysis.
• Investor relations. GenAI can also transform customer relationship management, including investor relations. The technology can be used to better target potential investors and to maintain ongoing relationships. Generating marketing materials, investor presentations and answers to investor queries can all be streamlined using GenAI. Investor chatbots can be deployed, giving investors the ability to have their questions answered more efficiently.
• Business support. HR, IT and legal are typical support functions within real estate, and GenAI can significantly impact all these areas. GenAI can be used in the hiring process by creating job specifications, screening resumes, performing background checks and supporting the interview process. Employee chatbots can streamline HR processes. AI can be used to develop and deploy code for new capabilities and to manage cybersecurity. Typical legal functions, such as contracting and document discovery, can be transformed. Procurement processes, such as vendor selection, bid analysis, purchase order generation and invoicing, can be enhanced.
• Asset management. GenAI can give asset managers the tools necessary to collect and analyze property-level data more effectively. This will lead to enhanced budgeting and forecasting. There is also the potential for GenAI to be used in leasing, ESG reporting, capital planning and risk identification. Reporting and scenario planning can be streamlined. In addition, specific to real estate investment trusts (REITs) and real estate fund organizations, asset managers can use GenAI to enhance functions like trade processing, performance management, fund accounting and administration. They can also automate operational tasks, like tracking and coding of investment management agreements and other compliance activities, as further outlined in the EY article, “Five priorities for winning with GenAI in wealth and asset management.”
• Finance and accounting. Generation of financial reports and forecasts, risk assessment and compliance, and fraud detection can all be streamlined with the help of GenAI. Tasks such as invoice generation and processing, payments, and billing have all been automated to some extent but can be further enhanced with AI.
• Property operations. AI is already being implemented to better manage property operations, such as energy management. GenAI can take this even further with much more dynamic energy optimization. Security and access control can be further enhanced. Tenant chatbots are becoming more pervasive, making processes, such as maintenance requests, account information and rent collection, even more efficient. Reporting can be more robust. Marketing and leasing processes can be significantly enhanced. Generation of marketing materials, tenant acquisition, leasing, and creation of presentations and reports will be impacted by GenAI tools.

AI implementation

Given the wide range of possibilities, real estate companies should start with developing a GenAI approach that includes:

1. Use case selection and process transformation. This should include company-specific immediate, medium- and long-term applications. Leveraging GenAI is about rethinking how existing activities should be performed. It’s a different mindset — a transformative mindset. A business case should be made as to how GenAI can help the business and what those benefits should be.
2. Technology roadmap and selection. Conduct “buy/build/acquire/wait” analysis to develop a model and acquisition plan for GenAI implementation. Evaluate the infrastructure that will be needed and create a report with estimated costs and timelines.
3. Responsible and ethical AI. Organizations must be purposeful and start with robust, responsible AI that performs as promised. First and foremost: train, educate and engage the workforce. Employees need a deeper understanding of AI leading practices and how the organization is using data. To replace uncertainty with confidence, employees should know how to use AI securely, responsibly and ethically. They also want to know more about what the organization itself is doing, including transparency about AI and data use, as well as having third parties review AI applications. It’s important for leadership to work with their risk, compliance and legal teams, as well as teams with experience developing digital policies and procedures, to create guidelines to inform their GenAI approach. Develop a scalable AI governance framework and continuation monitoring processes. Engage in dialogue and conduct scenario planning to mitigate risk and create a plan that puts the organization in the best position to succeed.
4. Organizational transformation roadmap. Build a plan for what GenAI should look like as it is being implemented and at regular intervals going forward. Formalize both an internal adoption program and an organizational transformation roadmap to outline goals and maintain accountability.
5. Talent transformation. GenAI adoption will require current employees to upskill themselves on new technology and may require hiring new talent. It will also have an impact on future resource count, especially in areas where GenAI has the most potential to introduce transformation. Aligning people strategy with business strategy is important to drive enterprise transformation.

As part of the technology roadmap and selection process, there is now a proliferation of tools available in the marketplace to start implementing GenAI capabilities. In addition to a cloud infrastructure, key foundational elements of a technology stack required for GenAI include:

1. Foundation models. These models are trained on a broad set of unlabeled data that can be used for different tasks. Real estate companies will need to build capabilities on top of a foundation model.
2. Data storage and retrieval. To build GenAI applications, real estate companies will need to build capabilities, such as a semantic layer, to efficiently store and retrieve both structured and unstructured contextual data. Most of the real estate use cases mentioned above rely on internal and external data.
3. Models and applications. Fine-tuned models and applications will need to be developed for specific use cases. There are application frameworks available that can help to accelerate the development and deployment of models. GenAI tools that help to enable some of the use cases described above have already begun to be developed and will improve over time.
4. Hosting options. There are two options to host foundational models: third-party hosting or self-hosting. Which option an organization uses will determine the encryption and security levels for the models, prompts and data.