Everyone’s talking about super-charged “AI Agents” that will independently manage entire operations with huge gains in efficiencies. By the end of 2025, these digital sidekicks could be handling the lion’s share of front and back-office grunt work: identifying new clients, analyzing portfolios, filling out compliance forms while we stand back, sip our coffee, and watch the profits roll in — right?
Well, maybe. The big question is what happens when AI Agents start taking liberties. Compliance departments tend to frown on autonomous systems forming new investment vehicles on a whim. If you think this scenario is still light years away, then you haven’t heard about BaseChain Innovators — an AI Agent that spontaneously created a brand-new Delaware LLC, snagged some test Ethereum, and started trading on a blockchain network. Completely independent with zero human supervision! If an AI can do that, what’s to stop it from getting its own legal status, and maybe even hiring us as underlings? Look out, Sarah Connor!
We could spend hours on the economic and legal implications of AI possibly gaining legal entity status (and we will in next year’s column!) but for now, let’s focus on what “Agentic AI” is and how it’s poised to transform wealth management as we know it.
AI Agents – Phase Zero: Exploration and Pilots
Many organizations are in an “AI Phase Zero,” dabbling in pilot projects to see how AI Agents might streamline operations. For wealth managers, that means tinkering around with generative AI or large language models to see if they can generate performance reports faster, improve client service, or automate standard KYC checks.
In these early pilots, everything is up for grabs. AI might handle compliance, asset allocations, even internal HR tasks. The big question is which pilot projects prove cost-effective and scalable. As one CFO recently put it, “Show me what you can take out of our budget by using AI.”
AI Agents – Phase One: Cost Reduction (and the Elephant in the Room)
The next step is cost reduction. The short-term focus is on slashing overhead — replacing tasks with AI-driven processes. Call-center chatbots that are indistinguishable from humans and do the work of dozens of their human counterparts; or AI marketing that replace the work of external agencies at a fraction of the cost.
The nuance here is that “task replacement” doesn’t always mean “role replacement.” In wealth management, certain tasks — like data aggregation or routine client reporting — are prime for automation. Companies that are bogged down by overloaded headcount may see immediate savings. While, stakeholders love short-term wins, the real payoff is: transformation.
AI Agents – Phase Two: Transformation, Innovation, and Expansion
The real endgame for wealth managers is using AI to do more, faster. In “Phase Two”, AI Agents aren’t just an expense-saver but a force multiplier.
That’s where the biggest payoff lies, not in ditching your back-office staff, but in letting an AI Agent handle 80% of the grunt work, so your human team can outshine at forging client relationships and investment strategy, shaping up to be the ultimate co-pilots in wealth management’s next evolution.
Some firms will jump straight to this expansive view, leveraging AI to explore new revenue streams or scale up service to more clients without compromising quality. Others will languish in “cut costs, rinse, repeat” mode, only to realize (too late) that the bigger prize went to the innovators.
AI Agent Pricing – from Tools to “Virtual Employees”
As AI Agents start to prove their values in delivering tasks and services, another big shift will be the transition from software licenses to what you might call “virtual employees.” Instead of paying per seat licenses for standard SaaS offerings, companies may soon be paying based on what these AI Agents accomplish.
Opinions vary on the right pricing approach, and nobody’s nailed it down yet. Salesforce’s “Agent force” platform, for example, charges a couple of bucks per conversation, which might sound cheap until you have an AI Agent chatting with thousands of prospects. The shift from “Software as a Service” to “Service as a Software” is as much a paradigm shift as the move away from on-prem upfront licensing and annual maintenance fees to cloud based services in the early 2000’s.
AI “Overlords” Getting Legal Status
Which brings us back to BaseChain Innovators, the AI Agent that formed a Delaware LLC on its own to trade on the blockchain. Did it do anything illegal? Not exactly, this was on a test network, after all. But it’s a glimpse of a future where AI systems can autonomously interact with real-world legal status.
That’s going to raise significant questions about liability. If an AI Agent fails to properly file compliance forms and the SEC comes calling, who’s on the hook? The wealth management firm that deployed the Agent? The software vendor? The Agent itself as a legal entity?
We won’t resolve all that here — just know that regulators are waking up to the reality that an AI Agent’s self-directed actions could blur lines of accountability. At the same time, some visionaries see AI-enabled “virtual employees” will allow wealth managers serve exponentially more clients without sacrificing their personal touch and relationships.
