Reading Time: 2 minutes

The investment management industry is moving fast on artificial intelligence. The largest institutional firms now draft investment theses in days instead of weeks and run agents that monitor markets in real time. They are racing toward alpha generation trying to achieve an extra 1% on their portfolio performance.

Family offices are not doing this.

They are doing something more practical and measurable. 

They are using AI to mitigate the operational burden of running a family office, not to beat a benchmark or other institutional investors. To make the back office vanish so the front of the house can focus on the work that matters.

This is part one of a three-part series on AI adoption in the family office. 

The 2026 Citi Institute report surveyed principals and chief investment officers across four continents. The headline number is that 22 percent of family offices now use AI for operational tasks or investment analysis. A year earlier it was 13 percent. Almost double the growth, which is uncommon to see that type of trend in the family office space. 

The more interesting finding is what family offices refuse to use it for. 

The research found that they are not making investment decisions with it due to the unreliability. This isn't to say that in a year from now the technology will be at a place where family offices feel more confident about its ability to create stronger investment decisions. 

They are also not feeding sensitive financial data into consumer chatbots. Cybersecurity is a major component and fear of many of these families that are ultra private. The data they hold is not purely financial; it is medical, marital, and intellectual property. It is the private record of how a family lives. A data breach can be a generational burden and the reward does not feel worth the risk right now. 

Many of these family offices surveyed have focused on a strong IT governance approach. Because privacy is not a constraint on AI adoption. You buy the enterprise version, not the cheapest one. You anonymize first, or you keep the work behind your own firewall. You wait until the frontier model can operate inside your security perimeter. This is slower than what large institutions do but it is also more durable.

Institutions with scale can have a larger budget for experimentation, and extra margin to get AI wrong. Family offices have none of those luxuries. With a smaller team, a modest budget, and a single principal who is wearing multiple hats, it can be overwhelming to implement AI into the operating system of the family office.

But don’t get me wrong, the goal is not a family office without humans, it is one where humans are freed to do what only humans can do. Build relationships and make nuanced judgments.  

The prize is not a better return on your investment. It is allowing AI to help you operate your family office like a business with lean operational costs and precision on decision making. 

At Legacy and Succession, we help families build their Family Office and AI adoption is becoming a huge component of that process. So where do we start taking a bite of the elephant? What is some low hanging fruit that AI can help solve?

You start with paper.

Every family office drowns in documents. The 120 page private placement memorandum. The partnership agreement with terms buried on page 60. The K1 that arrives in March and makes no sense. Reading all of it is the job, and it consumes the time of the people that need to be focused on more important responsibilities. 

This is why document review is the most adopted use of AI in the family office. The value is immediate and the risk, handled correctly, is low. AI reads the 120 page memorandum in the time it takes to pour a coffee and tells you which 20 pages actually require your judgment. It extracts key terms from a contract and flags unusual clauses. It takes four fund documents in four formats and produces one standardized summary, so you compare managers rather than document layouts.

Over here we have proprietary estate, legal, and PPM software that utilizes AI to extract data in minutes and saves hours to do so. It’s safe, efficient, and honestly better at tracking down important areas within the documents that we may have skipped over. 

But it decides nothing. One principal in the Citi study said it best. AI is an assistant, not a decision maker, and human legal review remains essential.

Building confidence in AI is a common roadblock most family offices face. Typically, they start by checking everything the AI produces. They verify the accuracy of every output and gradually reduce their manual oversight as they learn to trust the specific capabilities of the technology.

You already own the documents that are burying your team. What would your best people do with the time if they were no longer the ones reading every page?

Best,

David Ortiz

Vice President | Legacy & Succession

Family Succession is a private dispatch exploring the structural mechanics of enduring family wealth. If you received this from a peer and would like to join the list, you can request access [here].

Our Books