MIT Technology Review names its 35 Innovators Under 35 for 2026
MIT Technology Review's newsletter The Download led its September 8 edition with the reveal of its 2026 Innovators Under 35 list: 35 people from around the world doing what the publication calls groundbreaking scientific and technical work. The honorees were chosen from 550 nominations, with 44 expert judges helping the editors evaluate finalists, and each works in one of four categories: biotechnology, AI, computing and robotics, or climate and energy.
The same edition carried a feature, bylined Stephen Ornes, on using spiderwebs as a source of environmental DNA (eDNA) to survey wildlife. Counting animals for conservation is normally slow and costly, but spiderwebs trap genetic material shed by the spiders themselves, their prey, and bio-detritus such as saliva and pollen from nearby plants and animals. According to the newsletter, studies found no passive survey tool that matched spiderwebs' ability to identify which vertebrates are present in an area.
The newsletter's must-read roundup linked out to several other AI and tech stories. Anthropic's Claude formalized a computer-verified proof of Fermat's last theorem in 11 days, producing 13 million lines of proof, according to a linked Nature piece. Mistral raised $3.5 billion, described as a European record and the biggest equity round for a private European tech company, with linked coverage noting the company is betting on open models while US rivals keep theirs closed and is shifting strategy toward AI infrastructure, a pivot to data centers and services that has also drawn criticism. Elsewhere in the roundup: a blacklisted Chinese company allegedly kept buying Nvidia's best AI chips by having its US subsidiary ship them to firms serving China from elsewhere; Deutsche Telekom, Orange, Vodafone and Telefonica are reportedly in talks to form a consortium building a satellite-to-mobile venture meant to rival Starlink; tech companies are exploring Patagonia as a site for large AI data centers, citing cool temperatures, abundant energy and new reforms; and Australia is weighing a law that would let users switch off social media algorithms and penalize platforms that refuse.
The newsletter's quote of the day came from Sen. John Kennedy (R-La.), who told NBC's Meet the Press that the people who built AI are smart but poor communicators. The edition closed with a teaser for a separate, more detailed MIT Technology Review analysis of AI's energy footprint, bylined James O'Donnell and Casey Crownhart, based on interviews with two dozen experts, evaluation of different AI systems and prompts, and review of reports and projections; the newsletter text does not itself state the analysis's findings.
Key facts
- MIT Technology Review named 35 people to its 2026 Innovators Under 35 list, selected from 550 nominations with the help of 44 expert judges, across four categories: biotechnology, AI, computing and robotics, and climate and energy.
- A separate item describes using spiderwebs as an environmental DNA source for wildlife surveys, citing studies that found no passive tool matched spiderwebs' ability to identify vertebrates in an area.
- Anthropic's Claude produced a computer-verified, 13-million-line proof of Fermat's last theorem in 11 days, per a linked Nature report.
- Mistral raised $3.5 billion, called a European record and the largest equity round ever for a private European tech company.
- Other linked items cover a blacklisted Chinese company allegedly rerouting Nvidia chip purchases through a US subsidiary, a Deutsche Telekom-Orange-Vodafone-Telefonica consortium exploring a Starlink rival, and Australia's proposed opt-out law for social media algorithms.
Why it matters
The Innovators Under 35 list is MIT Technology Review's annual attempt to flag which young researchers and builders are worth watching before their work becomes widely known, spanning biotech, AI, computing and robotics, and climate and energy. The same edition shows AI's reach widening on several fronts at once: a formal mathematics proof produced largely by an AI system, a record European funding round for an AI lab, and a forthcoming deep dive into how much energy the AI boom actually consumes.
Who it affects
Readers tracking early-career scientists and technologists get a curated shortlist rather than having to scan hundreds of nominees themselves. The roundup items touch AI hardware supply chains (chip export controls), AI funding and infrastructure strategy (Mistral), telecom operators weighing satellite connectivity, and social media users in Australia who could gain a legal right to turn off algorithmic feeds.
How to use it
The full 2026 Innovators Under 35 list, the spiderweb-eDNA feature, and the AI energy footprint analysis are each published in full on MIT Technology Review's site; this newsletter edition only summarizes and links to them, along with one-line pointers to stories reported by NYT, CNBC, Reuters, Bloomberg, BBC, the Economist, Gizmodo, Wired and New Scientist.
How solid is it
The Innovators Under 35 list and the spiderweb-eDNA piece are MIT Technology Review's own reporting, with the latter bylined to Stephen Ornes. The must-read items are one-line teasers crediting other outlets for the underlying reporting, so the newsletter text itself does not contain the full detail behind claims like the Fermat proof's verification method or the Mistral round's investors.
Risks and caveats
As a digest, the edition packs many unrelated items into short summaries; the source text does not give a valuation or investor list for Mistral's raise, does not specify how the Fermat proof was verified beyond calling it computer-verified, and does not state penalty amounts or a timeline for Australia's proposed social media law. The teased AI energy footprint analysis is described only by its methodology here, not its conclusions.
“The people who have developed AI are very, very smart, but they're high IQ, stupid people. They're terrible marketers.”
— Sen. John Kennedy (R-La.), speaking to NBC's Meet the Press