The “vast majority” of family offices are “still trying to sort out” AI implementation, according to Lauren Phillips, director of Family Enterprise Advisory Services for PwC US. Family offices are hesitating  because they fear disruption and that AI implementation “will slow things down,” said Marie Samuel, solution consultant for finance software provider Sage.

Too many family offices with “legacy infrastructure” are “falling well behind” what’s needed to properly implement AI protocols, added Warren Finkel, managing director, Northeast for IT and cybersecurity firm Omega Systems. 

But delaying implementing artificial intelligence software isn’t an option, speakers warned. “The AI you’re using now is the worst it will ever be,” said Murali Nadarajah, CIO for Eton Solutions, a software provider and headline sponsor of the conference. “In three months, the software will be better, and it will be even better six months later.”

As regularly noted by FWR, the wealth management sector, including family offices, continues to explore and implement AI use cases, all the while mindful of how fast change is taking place. 

Pain points and metrics
Proper training, patience and identifying pain points are critical first steps for implementation, speakers said. 

“AI takes time to train,” said Jesse George, managing partner at Shade Tree Advisors, a family office based in Saratoga Springs, New York. “It’s an iterative process and you can’t expect AI to be perfect right away. Start with high reward and low risk projects. Make sure the first applications are not for mission critical matters and distinguish between what are acceptable errors and what aren’t.”

Family offices need to ask open-ended questions to identify pain points, Alexandre Lin, CEO of Sumit, a family office accounting software vendor, noted. “You need to understand what the problems are to get to the right solution,” Lin said. Similarly, family offices must set metrics with efficiency targets that are achievable, said Meagan Askew, director of PwC’s US Data AI and Analytics practice. 

Keep humans in the loop
Artificial intelligence is still not one hundred per cent accurate, and speakers stressed the need for human involvement when implementing AI, what Peter Bixler, senior manager for PwC’s Family Enterprise Advisory Services called keeping “a human in the loop.”

“AI is good for the first draft, not the final draft,” as Nader Souri, global head of corporate and digital banking for BNY, the conference’s venue partner, put it.

Family offices were urged to seek out experienced and empathetic software partners. Vendors have to understand what family offices are looking for and recognize their anxiety when replacing a legacy system, said Erin Hulse, CEO of Deviate Consulting, a family office software consultant.

Data is king
The whole point of implementing AI, of course, is achieving desired results and benefits, chiefly efficiency and productivity gains. 

Document processing, data entry and analytics were frequently cited use cases. For example, AI has reduced the time Shade Tree Advisors spends entering custodial statements by over 80 per cent, according to George.

Data has fueled the AI phenomenon, BNY’s Nader noted, and other speakers stressed the importance of maintaining  the integrity of AI’s source of data.

“Good data in means good data out and bad data in means bad data out,” said Sarmiento. PwC’s Askew advocated for appointing a “data steward” to oversee  the flow of data into AI. Alex Lisle, chief technology officer of cybersecurity firm Reality Defender warned that “data infiltration” by cyber criminals has become a “prevalent problem” in 2025, one that he “doesn’t see slowing down.”

Risks
Transferring  unmediated data to an AI platform is like posting personal financial data information on Facebook, said Yash Tulsani, chief financial officer of cybersecurity and compliance firm CalypsoAI. Family offices implementing AI “have to think about the impact of data security,” he cautioned.

Additional risks in the AI era include cyber criminals using deepfakes and social engineering. Deepfakes, or hyper-realistic, fabricated videos, photos, or audio recordings created using artificial intelligence which can be used to make someone appear to say or do something they never did, have become “go to” weapons for cyber criminals, according to Lisle.

What’s more, AI has made social engineering, or using psychological influence to induce people into performing actions or divulging confidential information, “infinitely easier,” Lisle added. For example, family offices are seeing increased incidents of AI simulating an executive’s speech to order an employee to perform harmful actions, such as authorizing the release of funds to bad actors, he said.

New era
For better or worse, speakers made clear that artificial intelligence has ushered in a new era for family offices. “We will be the last generation to manage only humans,” said Brianna Ferguson, head of strategic partnerships for BNY. 

But will there be fewer humans working at all?

As AI reduces time employees have to spend on repeatable tasks, there will be a “cascading effect,” said George. “We have no intention of letting people go, but it will impact the pace of hiring.”

In addition to AI, the conference also featured a look inside family office portfolios as seen through Addepar’s dataset, presented by the  asset management platform’s investment associate Katie Zhou. 

The need for family offices to consider the role of crisis management and kidnapping and ransom insurance was highlighted by Gordon Coyle, CEO of commercial insurance brokerage firm The Coyle Group and Steve Ward, CEO of security firm Polaris Risk.

Other conference speakers included Kelly Guinan, director, Responsible AI for PwC US, Jack Arrix, director, strategic alliances for Sage and Walter Hicks, the firm’s industry principal.

Tech Demos
The Fintech Summit featured tech demos from:

-- EtonAI on delivering real-time, intelligent answers to family members and office staff by tapping into structured and unstructured data.

-- Masttro on how secure, integrated, AI-powered intelligence can automate workflows and streamline reporting. 

-- AgiLink on how secure bill payment and specialized cloud accounting incorporates approval workflow, document management and investment data integration.

-- Archway Group on using its platform for real-time accounting visibility, investment aggregation, and complex ownership structures.

-- Advyzon on how its Auria family wealth management platform’s single interface combines communication, documents, tasks, and workflows. 

-- Bill on how integrated workflows, fraud prevention, and customized controls streamline complex bill payments.

-- Sumit on how its platform, driven by feedback from family office clients and advisors, provides one-click, customizable consolidated reporting.