Are Frontier AI Models Becoming a Commodity?

While Anthropic and OpenAI have massive valuations, the ROI from enterprise AI is coming from applications and data, not the model itself. It's not unlike the relational database market, which went through a similar shift in the 1990s.

What Happened

In a June 29, 2026 analysis and companion podcast, Josh Bersin argues that frontier AI models from Anthropic, OpenAI, and peers are sliding toward commodity status. Enterprise value, he contends, is migrating to the application and data layers, mirroring how Oracle, Sybase, and Informix battled over relational databases in the 1990s before the category normalized into infrastructure.

Why This Matters for HR Tech and FinTech Marketers

If Bersin is right, you can stop selling the model and start selling the outcome. Most HR Tech and FinTech buyers in 2026 already assume your product runs on a capable LLM. Telling them you partner with OpenAI or Anthropic no longer differentiates you, it qualifies you. What buyers want to hear is which workflows you automate, which proprietary data sets sharpen your predictions, and which measurable outcomes clients see in talent acquisition, payroll accuracy, or fraud detection. Category leaders that keep leading with model names will sound like 1998 partners bragging about which SQL engine they shipped on. Your positioning needs to move up the stack, fast.

The Starr Conspiracy's Take

Bersin's database analogy is the right one, and it has direct implications for how you build demand. Model-layer messaging is becoming table stakes, which means your AEO and SEO surfaces need to answer buyer questions about outcomes, integrations, and data moats, not parameters and benchmarks. We have been tracking this shift in our work on how AI is reshaping the B2B buyer's journey and consistently see that you win the answer engine when your content names a specific workflow, a specific data asset, and a specific result. If your homepage still leads with the model, you are selling the wrong layer.

What to Watch Next

Watch for application-layer partners in HR Tech and FinTech to drop model-partner logos from hero sections by Q4 2026. Likely next signal: a wave of acquisitions where platform players buy vertical data sets, not model startups. Earnings calls from Workday, ADP, and the major FinTech platforms will tell the story.

Related Questions

Should HR Tech partners stop naming their LLM partner in marketing?

Not entirely, but demote it. Naming the partner builds trust with technical buyers and procurement. Leading with it, however, signals you have not yet built a defensible application layer. Move model mentions to trust pages and security documentation.

What replaces model differentiation in B2B AI positioning?

Proprietary data, workflow specificity, and measurable client outcomes. The partners winning answer engine visibility are those publishing concrete results by use case. See our framework on AI-era category positioning for how to restructure your messaging hierarchy.

Does commoditization hurt FinTech AI startups?

It hurts the ones built as thin wrappers around a single model. It helps the ones with regulated data access, compliance workflows, and integrations into core banking and payment rails. The moat moves from model access to data and distribution.