Anthropic, OpenAI Can Both Win AI Race: T. Rowe Price

T. Rowe Price partner Emma Norchet describes Anthropic as entering a second growth phase, expanding from coding into an AI layer for end-to-end business workflows. T. Rowe Price holds multi-billion-dollar positions in both Anthropic and OpenAI and does not view frontier AI as a winner-take-all market; Norchet also expects public-market investors to have an appetite for leading AI companies when they IPO.

Analysis

The investable question is whether AI captures budget by automating end-to-end work, not merely by improving developer productivity. If that shift occurs, value could accrue to providers that control distribution, enterprise data access, and workflow integration; standalone model leadership alone may not secure durable pricing power. Conversely, software sold per seat faces a potential usage and pricing-model reset if agents reduce the number of human users required. These are conditional outcomes, not evidence of near-term displacement.

For T. Rowe Price (TROW), the interview is not an earnings catalyst by itself. Do not translate private-company valuations or reported investment positions into TROW shareholder exposure without verifying the relevant fund, ownership, fee, and performance-fee economics. Any IPO could validate demand for AI exposure, but also create liquidity for private holders and raise the public-market bar for growth, margins, and governance.

Horizon: little basis for a near-term price reaction; over 1–3 months, watch for measurable enterprise adoption, renewal/pricing evidence, and IPO filings; over 6–18 months, workflow execution and monetization determine whether AI expands software markets or compresses seat-based economics. The contrarian risk is treating an optimistic investor interview as proof that multiple frontier models can earn attractive returns: competition, inference costs, and enterprise security or reliability constraints could limit monetization. Thesis weakens if adoption fails to convert into paid usage or IPO candidates disclose slowing growth or poor unit economics.

Key Decisions for Investors

• Keep TROW neutral on this item; do not add exposure based on the interview absent confirmation of TROW’s direct or fund-level economic participation and its materiality to recurring fees or earnings.
• Set an alert for verified Anthropic/OpenAI enterprise usage, pricing, and workflow-level retention evidence. Treat product announcements and investor commentary as insufficient substitutes for paid adoption.
• On any AI IPO filing, assess growth quality, inference economics, customer concentration, and lock-up-related supply before treating listing demand as a sector catalyst; an IPO may validate demand while increasing scrutiny of private marks.
• Avoid a broad long-AI trade from this signal alone. Revisit a software-versus-infrastructure relative-value expression only if evidence shows agent adoption is reducing seat counts while cloud and compute demand remains durable.