Friday, March 1, 2024

In 2024, AI To Bring Small Improvements—and Bigger Threats -Dlight News

To get a sense of where advisor technology goes in 2024, Wealthmanagement.com spoke to a group of serial entrepreneurs in the space.

“The first thing we are going to see in 2024 is a lot more noise and buzz around AI, that is just the lay of the land right now,” said Mark Evans, president and CEO of Conquest Planning Inc. and a co-founder of NaviPlan. “The problem is you have to cut through all that buzz to get to the solutions,.”

Land and buzz are apt terms, given the parallels between the 185-year history of John Deere and the rise of artificial intelligence in advisor technology. 

Read More 2024 Industry Outlooks

While most associate the name with green and yellow tractors, Deere was a blacksmith who, in 1837, developed the first steel plow. A seemingly minor tweak to the cast-iron plows that existed at the time, the change led to profound improvements. Soil adhered to cast iron, gumming up the plows and forcing farmers to stop frequently to clean the machines. The soil didn’t stick to steel, letting farmers plow far more land and boosting the foundations for the modern agriculture industry. 

“He didn’t invent the plow, he manufactured a better plow, one made of steel that all the midwestern prairie soil didn’t stick too,” said Pete Giza, the co-founder of the early rebalancing platform RedBlack Software and the former chief product officer at at AdvisorPeak, which was sold to Addepar in 2021.

He predicts similar “incremental but meaningful” improvements using AI in 2024.

To be sure, “ninety-some odd percent of the startups around AI will fail,” he said. “There was no moat around it, and venture capitalists (threw) money at it because they didn’t know any better.”

But very real improvements in workflow efficiency and predictive analytics around client behavior are near, these entrepreneurs argue. For one, automation will help advisors bring better digital advice, and build relationships, with younger clients poised to inherit the wealth currently held by the Baby Boomers.

“Those younger clients won’t want to use the technology of their parents’ advisors,” Evans said. Successful AI ventures will have to avoid one of the failures that tripped up many of the early robo advisors, which rode a similar wave of hype, he said. Those robo advisors became their “own little island,” not connected to the rest of the digital ecosystem that advisors, and clients, are embedded in.

“Tools are available to feed (the clients), but what do you have to nurture them?” he asked. There are real opportunities for anticipatory, predictive technology that will help uncover more of a client’s assets and needs, he said.

Still, Evans worries the buzz around AI could lead to a backlash.

“Advisors are going to think, ‘This is great for me!’ but then find out they do not live up to the hype,” he said. “Will that dampen the interest for tools that take a little more time to put in place that will be beneficial?” 

But AI brings even bigger threats than advisor dissatisfaction.

And again, we can use the example of the success of the steel plow—and the unintended consequences—as a parallel.

So much of the prairie came under the plow that it virtually disappeared, leading in large part to widespread soil erosion, and ultimately, the Dust Bowl of the Great Depression.

Similarly, the widespread presence of AI rapidly embedded behind the scenes in almost all digital technology comes with very real threats. These entrepreneurs say cybersecurity and fraud could flourish in the wake. 

“Bigger organizations are typically okay, they have more resources but for the smaller RIAs, cybersecurity is going to be increasingly problematic,” Evans said.

“The cyber threats are going to get slicker,” he said. “You will stop seeing the grammar or spelling errors, that will go away soon with large language models able to fix text.” But that means for both advisors and the clients it will be all the easier to convince them with fake communications that perfectly mimic real messages and requests, and “suck them in to certain types of attacks,” said Evans.

Giza sees things similarly. “Now you are seeing really scary stuff,” he said, describing easily created deep fakes. Giza started in tech over three decades ago and has held positions at 3Com and Hewlett Packard, among others. 

“I was deep in security 20 years ago … and I hated it, having to wear a beeper 24-7″ to be on call for digital security breaches. AI is going to have a huge footprint in cybersecurity as well, he said.

“Antivirus programs are basically learning tools that are constantly updating the product, and now, instead of having to download a signature and update it, the AI is updating things in real time,” he said. That provides an opportunity for the AI bots of those with malicious intent to monitor the platforms in real time as well.

“What will be interesting is if AI is used to build social engineering environments, calling people and sending emails within a time period to get information from people before they realize what they’ve done,” said Giza.

“Unfortunately, the structure of the advisors out there are not set up to handle what is coming—it is up going to be up to their [technology and security] providers,” said Jay Jumper, president and CEO of SIGNiX, a cloud-based digital signature provider to many firms in the wealth management space.

In simplest terms this could be a lack of oversight—not necessarily external threats. He pointed to FINRA fining LPL $3 million for not adhering to Bulletin 2218‘s digital signature requirements to prevent fraud as an example.

But more extreme threats from bad actors are also on the rise. “If you talk about cyber, ransomware is just getting worse,” said Pem Guerry, another senior executive at SIGNiX.

“Firms and providers are trying to address the growth in these attacks with employee education but I think improvements in AI are only going to make the threats better and easier to implement.”

Related Articles

- Advertisement -

Latest Articles

- Advertisement -