Uber Burned Its Entire AI Budget in Four Months. Then It Cut the Team That Manages Hiring.

Two data points from Uber, separated by approximately six weeks, tell a complete story about where recruitment is heading in companies that have reached AI adoption maturity.

First: In May 2026, Uber's CTO disclosed that the company had exhausted its entire annual AI tools budget, primarily Anthropic Claude Code and Cursor, by April. Four months into the year. The company has since capped per-engineer AI spend at $1,500 per month and is treating the budget overrun as confirmation that AI tool adoption is running far ahead of initial projections.

Second: On 3 June 2026, Uber cut 23% of its People and Places division: the team responsible for HR, recruiting, and workplace operations.

These two events are not coincidentally timed. They are causally connected. Understanding the connection is critical for every recruitment agency with a tech-company client base.

What 95% AI Adoption Actually Means for Headcount

When Uber reports that 95% of its engineers use AI tools monthly and approximately 70% of committed code is AI-generated, the implication for the HR function is not immediately obvious. The connection runs through the headcount model.

Every company's HR and recruiting function exists to service a headcount model: a plan for how many people the company needs in which roles at which times. The HR function's size, how many recruiters, HR business partners, operations coordinators, is calibrated to the scale and pace of that headcount plan.

When AI tools double or triple individual engineer productivity, the headcount model changes. A team of 50 engineers using AI at high adoption rates produces the output that previously required 80-100 engineers. The growth that previously required 30 new engineering hires per year now requires 10-15. The recruiting infrastructure built for 30 hires per year is over-sized for 15.

This is the connection between Uber's AI budget and its HR cuts. The AI tools did not replace the HR team's work directly. They reduced the volume of work the HR team needed to do by reducing the headcount growth rate of the engineering function the HR team supported.

The COO's Question and What It Means for Agencies

When Uber COO Andrew Macdonald addressed the People and Places cuts, his framing was revealing: "If you're not actually able to draw a direct line to how many useful features and functionality you're shipping to your users, that trade becomes harder to justify."

This is the language of a technology product company applying product development logic to its HR function. The HR team is being evaluated on the same metric as a product team: are you shipping value that users can measure?

For Uber's internal HR team, the "users" are the engineering teams the HR function supports. The "features" are hires made, roles filled, processes improved, and team health maintained. When AI tools reduce the volume of hiring the engineering teams need, the metric the HR function ships against gets smaller, and the team size that is justified by that metric gets smaller with it.

Recruitment agencies servicing tech companies are being evaluated by the same logic, from the outside rather than the inside. The COO's question, "can you draw a direct line to features shipped for users," is what a CTO asks when reviewing whether to renew the agency relationship for tech-role recruiting.

The Headcount Model Your Clients Are Now Running

The Uber data points to a structural shift in how technology companies plan their headcount over the next 12-18 months. The pattern is visible across multiple companies at different stages of AI adoption:

• Early adoption (0-25% AI tool usage): Productivity gains are individual and variable. Headcount models do not yet change. Hiring continues at pre-AI rates.
• Scaling adoption (25-75% AI tool usage): Team-level productivity gains become visible in sprint velocity and output volume. Headcount growth slows. Hiring continues but fewer backfills are approved.
• Maturity adoption (75%+ AI tool usage): Organisation-level productivity gains are measurable. Headcount models are revised downward. The HR function's recruiting volume drops. The case for the same HR team size weakens.

Uber is at maturity (95% monthly usage). The 23% HR cut follows logically from the adoption data. The question for recruitment agencies is: where are your tech clients on this adoption curve, and what does that mean for the hiring volume you are projecting for them?

The "AI-Adjusted Headcount Model" That Recruitment Agencies Need to Offer

The conventional value proposition of a tech recruitment agency, "we will find you the engineers you need faster than you can find them yourself," assumes the client has a stable headcount plan and needs help executing against it.

At AI adoption maturity, the client's headcount plan itself is changing. The question shifts from "how do we hire 30 engineers this year" to "how many engineers do we actually need given our AI tool adoption rate, and what profile should they be."

The recruitment agencies that retain tech desk clients as AI adoption matures are the ones that can engage with the second question, not just the first. Specifically:

• AI-adjusted productivity modelling: Given the client's current AI tool adoption rate and the productivity data they have seen, what is the realistic headcount growth plan for the next 12 months? This is not a guess: it is a calculation the agency can help the client make based on adoption benchmarks from similar companies.
• Role-type pivot: At high AI adoption rates, the roles that have the highest hiring priority shift. The need for large numbers of mid-level engineers executing well-defined tasks reduces. The need for senior engineers who can direct AI systems, evaluate AI-generated code quality, and design architectures that AI tools can operate within increases. The agency that is already sourcing this profile is ahead of the conversation.
• HR function advisory: As the Uber example shows, the HR team itself is being evaluated on whether it justifies its cost in a reduced-volume hiring environment. Recruitment agencies that can help clients think about how to structure their HR function for an AI-adoption world, not just fill roles within the existing structure, are providing a service the HR team itself needs and cannot provide for itself.

The Companies That Will Cut HR Next

Uber is not unusual. It is early. The companies currently in the 25-75% AI tool adoption range, scaling adoption, where team-level productivity gains are becoming visible, will reach maturity adoption within 12-18 months. At that point, the same headcount model revision happens, and the same question about HR team size follows.

The recruitment agencies that serve those companies have a 12-18 month window to reposition the conversation from "help us hire" to "help us think about how we hire in an AI-assisted engineering environment." The agencies that wait for the Uber moment, when the HR budget conversation has already happened, will be responding to a changed brief rather than shaping it.

Uber burned its AI budget in four months. The HR team cut followed in six weeks. The sequence is predictable. The question is whether the recruitment agencies serving similar companies are already in front of the conversation or still behind it.