⏱️ Reading time: 8 min
Harvard announced on September 30, 2026 the opening of five AI professor searches within its Division of Science, according to a report from The Harvard Crimson.
📑 En este artículo
The number matters because it confirms that American universities, which until recently denied it privately, are willing to open several faculty positions at once to avoid losing the talent race against industry. The competition isn’t new, but the scale is: five simultaneous positions in a single division is a big bet, even for a university with Harvard’s endowment.
TL;DR
- Harvard opened 5 AI professor searches in its Division of Science on September 30, 2026.
- The call seeks researchers whose work bridges artificial intelligence with physics, chemistry, or biology.
- Universities are competing against tech industry salaries, which exceed the typical budget for a faculty position.
- No confirmed public budget: Harvard did not specify how much each new position will cost.
- The cluster hire process follows an academic calendar of joint committees and offers with shared compute.
What is a cluster hire of AI professors
A cluster hire of AI professors is the strategy a university uses to open several artificial intelligence faculty positions at once, coordinated across different departments, to recruit researchers whose work crosses AI with disciplines like physics, chemistry, or biology, sharing compute and budget instead of competing position by position.
What happened at Harvard
The statement reported by The Harvard Crimson on September 30 confirms five AI professor searches open within the university’s Division of Science. Harvard did not specify in the report the exact departments or the budget assigned to each position, a detail that the Crimson’s own story didn’t clarify either.
The exact profile they’re looking for isn’t known yet either: whether these are junior positions (assistant professor) or already-established faculty roles. What is public is the number, five, and the administrative framework: the division that groups physics, chemistry, mathematics, computer science, and earth sciences at Harvard.
Context and history: the competition for talent
The dispute between universities and industry over artificial intelligence faculty didn’t start in 2026. Over the past several years, labs like OpenAI, Google DeepMind, and Anthropic expanded their research teams with offers that include virtually unlimited compute, something no university department can match with public or endowment funds.
This phenomenon isn’t exclusive to Harvard. Over the past decade, many of the most-cited professors in deep learning took extended leave or resigned outright to join private labs, drawn by salaries and, above all, by access to GPU clusters that no university department can fund with ordinary endowment money.
That gap pushed several universities to rethink how they hire. Instead of opening one position at a time and waiting months between cycles, they group several searches under a single administrative umbrella. The goal is to reduce bureaucratic friction and offer, from the first contact, access to infrastructure shared across departments.
While Harvard announces this expansion, federal science agencies like the NSF are going through budget cuts that affect basic research in general. Harvard didn’t link both facts in its statement, but the timing coincidence raises the question of which universities can afford this strategy with their own funds and which can’t.
The mechanism behind the academic cluster hire
A cluster hire works differently from a traditional hiring process. Instead of each department posting its own opening and assembling its own committee, the division creates a joint committee that evaluates candidates for all five positions simultaneously. This allows comparing profiles across areas and spotting, for example, a physicist whose simulation work also interests the computer science department.
| Hiring model | When it’s used | Advantage | Limitation |
|---|---|---|---|
| Traditional search (one position) | A single opening in one department | Simple process, small committee | Slow cycles, limited compute offer |
| Cluster hire (several positions) | Strategic expansion in an area, like AI | Shares compute and attracts interdisciplinary candidates | Requires coordination between departments and a larger upfront budget |
The next step is the most expensive one. Each offer includes a startup package that normally covers postdoc salaries, lab space, and, increasingly, access to compute: dedicated GPUs or cloud credits to train and run inference on models. That line item has grown the most in recent AI hiring cycles, because a shared-use GPU cluster costs far more than traditional chemistry or biology lab equipment.
The full scheme looks like this:
flowchart TD
A["Division of Science identifies AI gap"] --> B["Joint committee across departments"]
B --> C["Searches opened in parallel"]
C --> D["Cross-committee interviews"]
D --> E["Offer with startup package"]
E --> F[("Access to shared compute")]
There’s no direct way to verify from outside how much each startup package costs in this specific case: Harvard doesn’t publish those figures due to salary confidentiality policy, and there’s no public record that breaks down spending by position.
Impact and analysis
Harvard’s bet matters beyond campus. If a university with Harvard’s financial endowment needs to open five searches at once to avoid falling behind, other public universities, with far tighter budgets, face an even bigger disadvantage against industry.
The underlying risk is young artificial intelligence faculty fleeing to private labs before they even publish their first independent paper. That empties out precisely the stretch of an academic career that trains the next generation of researchers, because a professor who never gets to direct a thesis leaves no students behind.
💭 Key point: A cluster hire doesn’t guarantee all five positions get filled at once: if not enough qualified candidates turn up, the university can leave positions open for the next cycle instead of lowering the bar.
On the other hand, the cluster hire has a positive side effect: it forces departments to share compute infrastructure that used to be purchased separately. That forced collaboration between physics, chemistry, and computer science can pay off in interdisciplinary research, even if the original motive was purely to compete for talent.
What’s next
The typical timeline for an academic search in the United States runs a full year. Applications are received between October and December, committees interview between January and March, and offers go out before the school year ends, with the new professor starting only in the following fall semester. Harvard hasn’t confirmed whether this process will follow that same timeline for the five positions.
What is predictable is that other universities will watch the outcome. If Harvard manages to fill all five positions with top-tier candidates, more science divisions are likely to replicate the joint-search model in 2027. If several positions stay vacant because eligible candidates prefer industry, that would also be a signal, and a worrying one, about the future of academic AI research.
Even if Harvard fills all five positions, recent history shows that retaining an AI researcher doesn’t end with signing the contract. Industry offers don’t stop once someone accepts an academic position, and several universities have had to renegotiate compute packages just a few years after the initial hire, to keep that same professor from accepting an outside offer.
If you work at a university and want to see what a call like this looks like, the full statement from Harvard’s Division of Science is quoted in the original story from The Harvard Crimson.
Frequently Asked Questions
What is a cluster hire of AI professors?
It’s the strategy of opening several artificial intelligence faculty positions at the same time, coordinated across departments, instead of filling each opening separately.
Why did Harvard open five searches at once?
The statement cited by The Harvard Crimson doesn’t detail the exact motivation, but it fits the general trend of universities competing as a bloc against industry’s salary and compute offers.
Which departments are involved in Harvard’s call?
Harvard didn’t specify the exact departments in the available report. The Division of Science groups physics, chemistry, mathematics, computer science, and earth sciences, so any of them could be included.
What does a startup package for an AI faculty member include?
Typically, postdoc salaries, lab space, and access to compute (dedicated GPUs or cloud credits), with the latter being the line item that grew the most in recent AI hiring cycles.
When would the new AI faculty start working at Harvard?
There’s no confirmed date. If the process follows the typical U.S. academic calendar, offers would go out before mid-2027 and the new hires would start in the following fall semester.
References
- The Harvard Crimson: original report on the five AI professor searches in the Division of Science, published on September 30, 2026.
- Harvard University: the university’s official site, with general information about its academic divisions.
- National Science Foundation (NSF): federal agency that funds much of the basic scientific research in the United States, whose budget is part of the context for this hiring push.
📱 Like this content? Follow @programacion on Telegram for daily tech content in Spanish: quick summaries, fresh content every day.
Featured image: Foto de Mikheil Kuzmidi en Unsplash
Did it work for you? Got a different error? Say so below: questions get answered and help the next reader.
Leave a comment
0 Comments