How xAI Stole a $7B Client from OpenAI by Showing Up

When Elon Musk's xAI started sending engineers to client offices, it wasn't announcing it. The company simply started showing up.

According to Bloomberg reporting published Friday, xAI has been deploying its technical staff directly to corporate sites — not to demo a product, but to do the work. Build integrations, connect data pipelines, customize the deployment. The sales pitch is not a slide deck. It is a proof of concept delivered on the client's infrastructure, by engineers who built the thing.

The strategy has already produced at least one concrete win. Shift4 Payments, a $7 billion payments processing company, plans to phase out OpenAI's ChatGPT in favor of xAI's Grok after an on-site engagement, CEO Taylor Lauber said in an interview with Bloomberg. The xAI team spent time inside Shift4's operations, building out the integration. That hands-on work closed the deal.

It is a notably old-fashioned way to sell enterprise software.

Consulting firms, but at scale

xAI is not the only company deploying engineers as a sales force. OpenAI launched its Frontier Alliance program in February — a partnership with BCG, McKinsey, Accenture, and Capgemini that pairs OpenAI's forward-deployed engineers with consulting teams to embed AI agents into enterprise workflows. The pitch is similar: we will not just hand you an API key, we will help you install this and make it work.

OpenAI's chief revenue officer Denise Dresser framed it as teaching companies to fish rather than fishing for them. "We do not want to build a model where we are doing the work," she told Reuters in February. "We want our customers to become self-sufficient." The consulting partners are the delivery mechanism — they have the bodies on the ground and the enterprise relationships.

Anthropic has taken a different route into the same problem. It is in discussions with private equity firms including Blackstone, Permira, and Hellman & Friedman to form a joint venture that would sell Claude technology across those firms' portfolio companies, per Reuters reporting this week. Anthropic is offering common equity stakes in the venture, betting that PE relationships will become a distribution moat.

OpenAI is pursuing a parallel track — also in talks with TPG, Advent, Bain, and Brookfield to form a separate joint venture with a pre-money valuation of roughly $10 billion and a PE commitment of around $4 billion, according to sources familiar with the matter. That venture would give PE firms preferred equity and board seats in exchange for steering portfolio companies toward OpenAI's enterprise tools.

The contrast in approach is revealing. xAI is building a direct services capability — engineers as salespeople, implementation as the differentiator. OpenAI is buying distribution through consulting firms and PE relationships. Anthropic is buying it through PE co-investment. All three paths lead to the same destination: getting human beings into enterprises to do the integration work that models alone cannot do.

Why implementation is the battleground

The shift makes sense when you look at why enterprise AI adoption has stalled. Most companies that tried to deploy AI at scale in the past two years ran into the same problem: the model is not the product. The product is the integration into existing workflows, the data pipelines that feed the model, the change management that gets employees to use it, and the observability systems that tell you whether it is working. None of that is solved by a better benchmark score.

"Companies have realized that siloed AI deployments do not deliver the value and they don't transform their company," Dresser said. xAI's approach is the logical extreme of that lesson — skip the proof of concept theater and just go do the implementation.

The enterprise AI market that OpenAI and Anthropic built was supposed to be software-as-a-service. The actual enterprise AI market, it turns out, is professional services. Shift4 did not switch from ChatGPT to Grok because Grok scored higher on a benchmark. xAI sent people to their office and solved a problem they actually had.

What this means for the competitive dynamics

The Shift4 win matters as a signal, not just a data point. Payments processing is a competitive market for AI — Stripe, Adyen, and Block are all active AI buyers. If xAI can win a reference customer in a visible, well-documented implementation, it changes the sales conversation for every other enterprise deal in the pipeline.

For OpenAI, the PE partnerships are an acknowledgment that direct sales at this scale is capital-intensive and slow. Buying distribution through Bain and TPG is expensive, but those firms together have visibility into hundreds of portfolio companies. The math only works if the PE relationship converts to long-term enterprise contracts rather than one-time deployments.

For Anthropic, which is widely considered ahead of OpenAI in enterprise penetration, the PE strategy is defensive — locking in relationships before xAI's field army can compete in the same accounts. Anthropic's enterprise annualized revenue reached $10 billion out of a total of $14 billion, per reporting in February. Defending that base while building a PE distribution channel is a different game than xAI's direct approach.

The race to own enterprise AI is becoming a race to own the implementation layer. And in that race, the engineers are the salespeople.