Large RIAs Name ChatGPT, Copilot and Claude More Often Than Wealth AI Tools

For large independent RIAs, AI is showing up in hiring plans well before advisor-specific products dominate the picture. Job posts more often name ChatGPT, Copilot and Claude than tools built for wealth management, while the share of firms mentioning AI drops sharply below $25 billion in assets. The gap points to uneven adoption across the industry and leaves a harder question: whether hiring for AI skills will translate into lower operating costs.

Caddi’s research found that 39 of 201 independent RIAs posting jobs in the third quarter of 2026 named at least one AI tool. That was 19.4%, compared with 1.1% two years earlier. Across the broader study period, from late 2023, 76 of the 455 firms examined had posted at least one job mentioning AI.

The figures capture what firms say they want in job candidates, rather than how extensively employees use the tools. Caddi reviewed job postings and current employees’ LinkedIn profiles, counting posts that named an AI product or asked for AI skills. The study covered the 455 largest independent wealth management RIAs, ranked by assets reported on Form ADV.

General-purpose tools lead the hiring signal

Across the 12 months through September 2026, Copilot appeared in job posts at 34 firms, ChatGPT at 32 and Claude at 27. Forty-one firms mentioned AI without naming a product. OpenAI’s GPT models were named directly by eight firms. Zocks, an advisor-focused meeting notetaker, was the only wealth-specific AI tool among the most frequently named products, appearing at seven firms—the same count as Google’s Gemini.

That distribution suggests hiring teams are asking for experience with broadly available AI systems more often than familiarity with specialized advisor software. It doesn’t establish which tools firms have put into daily use, or whether the named products are intended for client work, internal operations or both. A job post can signal interest and skill requirements, but it offers limited evidence about actual deployment or results.

Spending may be more widespread than the job-posting figures alone suggest. In a separate Caddi survey of 17 operations leaders at RIAs with more than $1 billion in assets, 53% said their firm pays for an advisor notetaker such as Jump, Zocks or Zeplyn. That result covers a small group of operations leaders and a specific product category; it shouldn’t be read as an adoption rate for all RIAs or all AI tools.

Adoption signals cluster among the largest firms

Firm size is a clear dividing line in the job data. Eight of the 10 largest independent RIAs had posted at least one job naming an AI tool. The share was 63.6% among firms with $25 billion to $100 billion in assets, then fell to 26.5% for those with $10 billion to $25 billion. About one in 10 firms in the $2.5 billion-to-$10 billion range had such a post; none below $2.5 billion did.

The pattern also appears in hiring for people who can build on large language models. About 21% of firms with $25 billion to $100 billion in assets posted jobs mentioning related tools or skills. That share dropped to 7.4% among firms with $10 billion to $25 billion and was close to zero below that range. Larger firms may have more scope to hire for internal development, but these figures don’t show whether they have completed projects or achieved measurable savings.

“The largest RIAs have stopped asking whether to use AI. The question for 2027 is whether anyone can show it changed the cost of running the firm, and right now very few can,” said Alejandro Castellano, Caddi’s co-founder and CEO. His comment frames the distinction between interest and business impact: postings can reveal a shift in hiring priorities, while cost outcomes require evidence the job data doesn’t provide.

Product boundaries are shifting

The divide between general-purpose AI and wealth-specific tools may become less clear. In September, Anthropic introduced Claude for Financial Advisors, with connections to Schwab, Orion, Wealth.com, Zocks and other companies serving wealth management. That development puts a general-purpose AI provider closer to advisor workflows, though the existence of connections alone doesn’t establish how widely firms will use them or what operational effects they’ll have.

Other research offers a useful comparison. Recent Schwab survey findings put advisor use of agentic AI—systems that make decisions and take actions within workflows—at 7%. In the same study, 26% said they want job candidates comfortable with AI tools. Schwab’s 2026 RIA Benchmarking Study also found that nearly half of surveyed RIAs had used AI to write job descriptions, while more than a third used it to develop interview questions.

Together, the findings point to a market where AI is entering hiring, recruiting and selected workflows, but adoption varies by firm size and task. For shareholders, employees and clients, the next meaningful measure will be whether firms can connect those uses to changes in operating costs, staff capacity or service. The current job-posting evidence shows growing demand for AI skills; it doesn’t yet answer that performance question.