AI Buildout Tops $1.5T, Advisors Expected to Stay

A May social event in Orange County hosted by Women in ETFs and the CFA Society of Orange County brought industry professionals together to discuss AI’s impact on finance. The session was moderated by Jane Edmondson, head of index product strategy at TMX VettaFi and co-head of Women in ETFs’ Southern California chapter. Panelists were Julia Bonafede, CFA, president of JobsPeak Advisors; Lu Yu, CFA, managing director and lead portfolio manager at Virtus Systematic; and Jon Cook, founder of Keynote Content.

Edmondson opened by stating that since ChatGPT’s late-2022 debut roughly $1.5 trillion has been directed toward AI deployment. She compared that sum to the projected U.S. defense budget for 2027 and said major technology firms are raising capital expenditures and tapping cash and debt to fund AI projects. Edmondson added that about 49,000 layoffs have been attributed to AI integration so far this year.

Panelists described current applications of AI in investment work. Lu Yu said she developed an AI-based factor in 2020 inspired by Google’s “Attention Is All You Need” paper and has used AI tools to process unstructured data. She reported that AI sentiment analysis flagged weakening performance in Russian stocks before Russia’s 2022 invasion of Ukraine, which allowed her team to exit positions and reallocate energy exposure to companies such as Petrobras in Brazil.

Julia Bonafede discussed limits of current systems. An early adopter of deep reinforcement learning, she warned that model performance can decay and that poor-quality input data can produce misleading outputs. Bonafede called for a broad retraining effort for finance professionals to learn how to validate AI results and maintain oversight of models.

Jon Cook framed AI as an efficiency tool for advisors. He described the technology as an “intelligent data mesh” that can automate manual tasks such as client-relationship-management updates and note-taking, freeing advisors to focus on client-facing work like legacy planning and crisis response. Cook said removing advisors from technology workflows harms client outcomes.

Panelists agreed that AI will automate routine tasks while increasing demand for human oversight and financial judgment. They said prompt-engineering skills will be useful but must be paired with domain expertise to test and validate AI-generated recommendations. Lu Yu said, “I still believe our capability to think, to have our own critical thinking, to be creative, will have us ahead of all the machines that we created.”