Available evidence indicates that, although the western U.S. megadrought has continued and water scarcity has worsened, governments have not newly begun using eminent domain or equivalent legal tools since mid‑2021 to broadly seize or revoke privately held water rights because of that scarcity.

Key points:

• Water scarcity has clearly intensified. The Southwest remains in a multi‑decade megadrought, the driest such period in at least 1,200 years, and Colorado River shortages have triggered repeated mandatory cuts for Arizona, Nevada and Mexico, administered through interstate compacts and Bureau of Reclamation shortage rules—not through seizure of private rights. (en.wikipedia.org)
• Western water law already treats water as publicly owned with usufructuary rights. In states like Colorado, all surface and groundwater are public resources; private parties hold a right to use water subject to prior‑appropriation priority and availability, not absolute ownership. (law.justia.com) Likewise, California’s public‑trust doctrine (e.g., the Mono Lake decision) has long allowed the state to re‑evaluate and adjust even previously granted rights to protect public trust uses. (en.wikipedia.org) These doctrines predate the 2021 podcast and are not new drought‑era seizures.
• Eminent‑domain authority over water‑related assets exists and has been used historically, but not newly for drought‑driven reallocation of private rights. For example, Grand Junction, Colorado condemned portions of ranchers’ water rights on Kannah Creek for municipal supply in 1911, a use later upheld in City of Grand Junction v. Kannah Creek Ass’n and discussed in contemporary water‑education materials. (law.justia.com) There have also been eminent‑domain efforts involving water utilities and distribution systems (e.g., California cities seeking to condemn investor‑owned water utilities), framed as service/cost issues, not as climate‑emergency takings of farmers’ or other private appropriators’ rights. (nossaman.com) These examples are limited and mostly historical or utility‑focused, not the broad “sorry, need it back” reallocation Chamath envisioned.
• Recent drought responses have relied on regulation and voluntary deals, not condemnation of water rights.
  • California’s Sustainable Groundwater Management Act (2014) is being enforced more aggressively—for example, placing the Tulare Lake groundwater subbasin on “probation,” adding pumping fees and state oversight—but this is regulatory management authority enacted years before 2021, not eminent‑domain seizure of vested rights. (en.wikipedia.org)
  • Colorado River cutbacks and conservation programs emphasize voluntary, compensated reductions (fallowing, system‑conservation agreements, water banking, tribal leasing), and states have explicitly sought to avoid mandatory federal reallocations that would spark litigation. (latimes.com)
• Several western statutes actually limit eminent domain over active water rights. California’s Water Code §60230 authorizes condemnation of property for groundwater replenishment but forbids using eminent domain to take “water and water rights already devoted to beneficial use.” (law.justia.com) The California Constitution similarly bars use of eminent domain to acquire Delta water rights or contracts for export. (law.justia.com) Colorado law has long debated restricting condemnation of water rights, and its water statutes emphasize that state agencies are not thereby authorized to acquire water by eminent domain. (landreport.com) These trends cut against a new wave of eminent‑domain takings of private rights.

Taken together, this suggests: (1) the conditions Chamath worried about (severe western water scarcity) have materialized or worsened, but (2) as of late 2025, governments have not in practice begun broadly seizing or cancelling privately held water rights via eminent domain or functionally equivalent legal mechanisms in response. Nor have there been high‑profile, post‑2021 cases that match his scenario of governments simply “taking back” water from private right‑holders at scale.

Because the prediction was open‑ended (“eventually”) and concerns a structural shift that may yet occur in the future, it’s too early to classify it as definitively wrong. At the same time, there is no solid evidence that it has already come true. Hence the most accurate assessment is inconclusive (too early) rather than right or wrong.