← All writing
Sales Jun 17, 2026 5 min read

The companies with the best AI on earth are hiring salespeople faster than SaaS is.

Christopher Dorsey

Christopher Dorsey

AI & MadTech Advisor · Enterprise Sales Leader

TL;DR

At OpenAI and Anthropic, go-to-market is the single largest hiring category: roughly one in five open roles, more than any other department, and a higher share than a typical SaaS company runs. The firms most able to automate their own sales are staffing it harder than anyone. The shape of the deal explains which sales jobs are safe: once a purchase involves multiple stakeholders, a novel budget, and procurement, the work goes back to humans. Aim your career at the ambiguous, trust-heavy, category-creating motions and away from the transactional ones.

If you want to know whether AI is coming for sales jobs, you can read another think piece, or you can look at what the two companies building the most capable AI on the planet do with their own hiring. At both OpenAI and Anthropic, go-to-market is the largest single category of open roles, around one in five, ahead of research and ahead of engineering. The businesses best positioned to replace their own salespeople with their own models are hiring salespeople faster than ordinary SaaS does.

That's the strongest evidence in the whole “will AI replace sellers” argument, and it points the other way. These are not companies that lack the technology or the nerve to automate. If a frontier lab could replace its sales org with its own model, it would do it loudly, as a case study. Instead they're building human go-to-market teams as fast as they can hire them.

Why the hardest AI sale still needs a human

The reason is in the shape of the deal. Once average contract value crosses the line where a purchase involves several stakeholders, a budget nobody planned for, and a procurement cycle, you need a human in the room. Frontier AI is the hardest thing in tech to sell right now: high trust, high ambiguity, a category the buyer didn't have last year, change management baked into every rollout, and a price tag that draws a crowd of approvers. Those are the conditions under which a human seller still beats any automation, including the seller's own product.

Their own hiring tells you which sales work lasts. The transactional, single-threaded, self-serve motion, where a buyer could finish the purchase without ever speaking to a person, is the part their products will absorb. The seat that survives is the one selling ambiguity into the enterprise.

I've been the first seller in the building twice, carrying products that reached the market before a budget line existed for them. That kind of selling is the work the data now points to as lasting: high-trust, category-creating, human-heavy, where most of the deciding happens in rooms you're not in and trust is the only thing speaking for you.

Audit your own motion

If you carry a number, look hard at how much judgment and trust your motion really requires. If your day is mostly volume and your deals close without much human persuasion, that's the part most exposed to automation, and it's worth moving toward complexity on purpose. If you build or lead a sales org, concentrate your human firepower where ambiguity is highest and let the tooling handle the transactional tail. At the top of the market, this job concentrates into fewer, more valuable seats. That's where the sellers worth hiring will be.

Share this post

About the author

Christopher Dorsey

Christopher Dorsey

Enterprise Sales Leader · AI Go-To-Market · Startup Advisor · Denver, CO

Fifteen years selling technology to Fortune 500 brands across AI, advertising, and data infrastructure — most recently at Zeta Global, Oracle, and Fastly. Currently advising founders and sales leaders on AI go-to-market and Generative Engine Optimization.

Questions, pushback, or just want to compare notes?