Palantir on steroids: How AI labs are taking over the consulting model

The model that OpenAI and Anthropic are adopting is not entirely new. The pioneer of the "forward-deployed engineers" concept was Palantir, which for years has been sending its technical teams directly to clients to build custom solutions on its platform. OpenAI and Anthropic are now scaling this approach to a level Palantir could never have dreamed of.

Instead of closing individual enterprise contracts one by one, by partnering with private equity funds they gain access to thousands of companies at once. Each of the partners — TPG, Blackstone, Goldman Sachs, and Bain Capital — manages a vast portfolio of companies across the mid-market economy. It is into these companies that teams of AI engineers will now flow, implementing GPT and Claude models directly into business processes.

According to Crypto Briefing, which reported on the deals on May 13, 2026, this is a strategy that resembles McKinsey with GPUs more than a classic AI lab.

Two deals, two completely different risk profiles

An interesting contrast emerges in how the two companies approached these joint ventures. One of the deals — according to available information, likely OpenAI's — includes a guaranteed annual return of 17.5% for investors. That is a bold commitment from a company that, by its own projections, will post a loss of around $14 billion this year.

Anthropic, by contrast, offers no such guarantee. This difference reflects the distinct DNA of the two companies: OpenAI bets on aggressive growth and is willing to take on greater financial risk, while Anthropic proceeds more conservatively. For investors, this means a clear choice — a more certain return with OpenAI, or a potentially higher but unguaranteed return with Anthropic.

Goldman Sachs as the key to fintech

Goldman Sachs's involvement is particularly significant for Anthropic. The partnership gives the company direct access to fintech verticals, specifically in wealth management, lending, and insurance. This fits perfectly into Anthropic's broader strategy, as the company previously launched specialized financial agents for banks, funds, and insurance companies.

For comparison: OpenAI is meanwhile already testing integration of ChatGPT with users' bank accounts via Plaid and offers an AI financial advisor as part of ChatGPT Pro for $200 per month. Both companies are thus targeting the same market, but each via a different route — Anthropic through institutional partnerships, OpenAI through a direct relationship with the end user.

Why consulting firms should be worried

Until now, the model for AI implementation in companies worked like this: consulting firms such as Accenture, Deloitte, or BCG would purchase AI model licenses from OpenAI or Anthropic and then implement them for their clients at high hourly rates. OpenAI and Anthropic are now cutting out this middleman.

By deploying their own engineers directly to clients, they capture both technology revenue and service revenue. This is a classic vertically integrated model, where the technology vendor also provides the implementation. For consulting firms, this represents an existential threat — and the irony is that Bain Capital, one of the partners in these joint ventures, has its own consulting division, Bain & Company, which is directly threatened by this model.

The numbers that explain the rush

Why are OpenAI and Anthropic so aggressively entering consulting? The answer lies in the data. According to an analysis by Menlo Ventures, OpenAI's share of the enterprise LLM API market dropped from 50% to roughly 25% between late 2023 and mid-2025. Customers are increasingly willing to switch between models depending on which one is leading the benchmarks.

Services, by contrast, create a much stickier customer relationship. Once your engineers are deeply integrated into a company's processes, the switching costs to a competitor rise dramatically. An internal memo from OpenAI CFO Denise Dresser from April 2026, obtained by The Verge, sums it up unequivocally: "Adoption of multiple products makes us hard to replace."

What this means for Czech companies

Czech companies will, for now, feel these global trends more indirectly, but the impact on the local market will come faster than it might seem. Consulting branches of large firms like Accenture or Deloitte have a significant presence in the Czech Republic, and their business model based on AI implementation will come under increasing pressure.

At the same time, this means that Czech companies should start building their own AI competencies instead of relying exclusively on external consultants. If AI model implementation becomes a "commodity" delivered directly by the vendor, the role of the middleman will radically transform.

Also relevant for the Czech market is the fact that Anthropic's Claude supports Czech at a very good level and ChatGPT has been available in Czech for years. Companies thus have direct access to both technologies on which these joint ventures are built.

Broader context: The war for the enterprise market escalates

The joint ventures with private equity funds are just one of many fronts on which OpenAI and Anthropic are clashing. OpenAI recently partnered with Amazon Bedrock to expand the distribution of its models beyond the Microsoft ecosystem. Anthropic, in turn, launched Claude for lawyers with over 20 connectors and integration with Microsoft 365 and introduced financial agents targeting the banking sector.

Both companies, according to market information, are preparing for an IPO as early as 2026, which makes their push to build stable and predictable revenue streams even more logical. After all, public market investors love recurring service revenue far more than volatile API revenue.