Students looking for a paid summer in a lab still start with the letters REU. In 2025 that door got smaller. AIP’s FYI service reported on 2 May 2025 that funding cuts were hitting science career pipelines. Fifteen active Research Experiences for Undergraduates grants were terminated in April 2025. The Graduate Research Fellowship Program made 1,000 awards in 2025, against more than 2,000 in prior years. Those are the counts in that article. A separate claim that only about 52 REU grants remained, against about 200, came from researchers’ social-media tallies and was reported secondhand. This post does not use that tally. The terminated-grant count and the fellowship count are enough.
An REU site is a funded cohort. When the grant stops, the cohort often stops with it, even if the science in the lab continues. A smaller GRFP class is a graduate-school count. Undergraduate students meet it later, when they apply for fellowships, which is why it belongs in the same conversation as a summer lab. A summer plan that assumes “the national list will have a spot” is using a list that just lost sites.
Where a summer can still sit
Department programs did not all depend on those fifteen grants. On 17 September 2026, @GeostatsGuy wrote about mentoring a rising senior through a department summer research program at a large public campus. That post is one professor’s account of one campus program. The post has no vacancy list. It does show a department running a summer cohort under its own name. Your version of the search is the undergraduate research office, the department’s own summer page, and the faculty who already teach you.
Professional societies run programs too. On 15 July 2026, @TheSIAMNews pointed at an eight-week undergraduate summer research program for 2027. Read the society’s own page before you trust a repost, and read the eligibility line. A society program is a third door beside a federal site and a department program. Hospital and lab summer programs in quantitative fields are a fourth, when they publish a date and a stipend. Aggregator accounts on social media are not a door. Confirm the deadline on the host’s site.
The practical order is local first. A professor who has seen your work can take one student without a national grant. A national site that lost its grant cannot. Spend the first week of the search on your own department. Spend the second on society and institute programs with a posted stipend and a posted date. Spend the third on whatever federal sites are still open, and read each site’s date yourself. There is no single national deadline.
How to read a program page
Write down four lines before you apply anywhere.
Who pays, and how much, if the page says. A program that is silent about pay may still be worth a question. It is not yet a plan you can budget.
Who is eligible. Class year, major, citizenship, and whether the program wants students from outside the host campus. A page you do not qualify for is a closed page.
The date. Site deadlines are not synchronized. A February date and an April date can both be real.
What you would do in week one. If the page names a method or a dataset, you can write a specific note. If it only says “research experience,” you can still apply, and your note should name a skill you have used.
Fifteen terminated grants do not tell you that your department’s program closed. Email the undergraduate chair and ask whether the summer program is running, who funds it, and when students are chosen. That email is allowed to be short. The AIP article is the reason to ask this year instead of assuming last year’s poster is still true.
A backup that is not a consolation prize
If the summer programs you wanted are gone, a term-time lab for credit or pay is the backup that still produces a letter and a skill. It will not look like an eight-week cohort. It can still be the thing you describe in a later application. Put it on the calendar as a real commitment, with a weekly hour and a named professor, or it will dissolve into office hours you meant to attend.
Do not inflate the AIP counts in a club presentation. Fifteen grants terminated in April 2025. One thousand GRFP awards in 2025, versus more than two thousand before. The “about 52 sites” figure stays off your slide unless you are ready to say it is a secondhand social-media tally. Precision is how you keep the bad news from becoming a rumor that stops a first-year student from writing the department.
How Flightpath helps
Faculty belong on the record when a professor has seen your work. You still ask them yourself. Flightpath can hold a draft of that note, and you approve the text before it is sent. The portal plan by year should show a summer research line in the year you can actually do it, including a department program that is not on a national list.
Pin a program only when it is a public page with a date. The tracker can hold those applications the way it holds jobs, from saved to applied. Gaps are against facts you confirmed, so a method you have used in class can support the note, and a method you have only read about cannot. Eligibility labels matter here too when a program restricts citizenship or class year. A role or program you cannot take is not ranked above one you can take. The AIP counts stay in the May 2025 article. The email to your department stays yours to send.