Amazon pledges $20M to the Colorado River without its own water data

Share
Amazon pledges $20M to the Colorado River without its own water data

Amazon put $20 million into the Colorado River this week — and still won't say how much water its data centers take out of it.

Amazon announced Monday that it will spend $20 million on water conservation across the Colorado River basin, as part of a new corporate coalition targeting $100 million over the next two years. The basin supplies 40 million people across the western United States and is in what federal forecasters call an unprecedented drought, with Lake Mead still declining. The money is real and the projects are specific: Amazon will fund leak detection and repairs with the Municipal Water District of Orange County, a $2.2 million effort over three years that the company says should conserve 189 million gallons a year, plus a $2.4 million partnership with agricultural technology firm Arable to give Native American farmers in the Lower Basin real-time irrigation data.

The coalition, called the Colorado River Basin Collaborative, includes Aldi and The Coca-Cola Foundation among its participants and has $32 million committed so far, according to the group's Monday release. Amazon chief sustainability officer Kara Hurst framed the model as an extension of the Climate Pledge: "We're bringing that same model to the Colorado River because water and climate are deeply connected." She added that the coalition will deploy funding into projects that are already vetted, starting now through 2028.

What Amazon did not provide is a baseline. Asked about the company's total water consumption, Hurst pointed instead to progress against Amazon's pledge to be water positive — returning more water than it uses — by the end of the decade, saying the company is 75 percent of the way there after returning 3 gallons for every 4 it used last year. In June the company disclosed that its data centers used 2.5 billion gallons worldwide, about 5 percent of what metro Seattle consumes annually, but that figure excludes the water consumed generating the electricity those facilities draw. An Amazon spokesperson told the Las Vegas Review-Journal that some of the new projects will be in Southern Nevada, but that the company is not ready to detail them — leaving the state with the most direct stake in Lake Mead's level with no sense of what the investment means for the region.

That gap is the story. A $20 million pledge against an undisclosed baseline cannot be audited, and a water-positive claim that sets aside the water embedded in power consumption answers a smaller question than the one communities along the river are asking. Data center cooling is only half of the industry's water footprint; the other half sits upstream in the generation mix, and it is the half most companies leave out of the number. Water policy is now moving on the same track as electricity policy — we covered the cooling side of this in August — AI data centers are turning to recycled wastewater to cool off — with regulators and utilities both starting to demand figures rather than pledges.

What to watch: whether Amazon names its Nevada projects, and whether the company publishes a basin-level water number that includes power generation. Until it does, the pledge is a donation, not disclosure.

Should a water-positive claim require a published total and third-party audit before it counts? Tell us in the comments.

Read more

Byteification turns Qwen3 and Llama 3 into byte-level models

Byteification turns Qwen3 and Llama 3 into byte-level models

The tokenizer is the last piece of the language-model pipeline that was never learned end to end — and a Nature paper out this week shows it can be swapped out for less than one percent of a pretraining budget. Researchers have retrofitted four open subword language models into true byte-level models using less than 1% of a typical pretraining budget, and the result — published 7 October in Nature — closes the performance gap that has kept byte-level models a niche curiosity for nearly a decad

OpenAI's dots go mobile and can start your Codex work

OpenAI's dots go mobile and can start your Codex work

One product update and two benchmark papers this hour, all circling the same question: how much of the work can you actually hand over, and what happens to the part you didn't? OpenAI's dots — the always-on agents that get their own cloud computer — can now be created and configured entirely from the ChatGPT app on iOS and Android, and a dot can start Codex work or pick up an existing Codex thread on your behalf. The October 9 release, billed as the first of a fresh run of updates for dots, mo

Coding agents as a hospital: Cockroach's five-month results

Coding agents as a hospital: Cockroach's five-month results

The agentic-coding debate keeps being framed as speed versus trust. Cockroach Labs just published the most detailed trust-first ledger yet — five months of numbers, including the parts that didn't work. Cockroach Labs ran its codebase like a teaching hospital for five months and says the pipeline merged well over a million lines of code with seven total reverts. In the setup the company calls MOLT Sinai, GitHub issues are patients, merging is discharge, and the humans in charge are Chiefs of M

Today in AI — October 10, 2026

Today in AI — October 10, 2026

A day where the boring machinery took center stage: proof checkers, order books, warehouses and the humble text message — plus fresh arguments about who gets to declare any of it safe. Models & Research * Mathematicians get a reliability primer for the Lean Theorem Prover. A guest post by Thomas Hales on Terence Tao's blog walks through what Lean actually guarantees — and devotes itself to the "Summer of Soundness Bugs," the string of kernel bugs found this summer that let false proofs thr