Dario Amodei says Anthropic plans for notably more than 10 GW of compute in 2027-2028 in a new Dwarkesh Patel interview. At 57:26:

Dwarkesh Patel: Suppose Anthropic's compute keeps 3x-ing a year, and then by 2027-2028 you have 10 GW. Multiply that by, as you say, $10bn. So then it's $100bn a year.
Dario Amodei: I don't want to give exact numbers for Anthropic, but these numbers are too small.

I think Feb 2026 is too early for full commitments for 2028 that can't be delayed, because most of the capex in a datacenter is in compute equipment, which likely only needs to be purchased within the last year before it goes online. If the revenue isn't there, the purchase can likely be delayed, possibly until the next generation of compute hardware.

The compute hardware is about 70% of the capex cost. For the other things (that mostly have to be built first), the useful life is 10-15 years rather than 5 years, and so the yearly commitments are disproportionately smaller than for compute, and there is no opex to speak of. This cuts $10-15bn per year per GW for a datacenter that's already online down to maybe $2bn per year for an empty shell of a datacenter (including power purchase commitments, which could take the form of on-site power, with similarly long useful life). Thus an empty shell of a datacenter (even a powered one) can wait for revenue that triggers the purchase of compute equipment at about 6x less in ongoing yearly payments, given a notice (about the need for a delay) of a few months to a year before the planned time of the datacenter going online. If even the empty shell wasn't built yet and only the land and permissions are being set up, delaying is even cheaper, and that can happen with 1-2 years of notice before the planned time of going online, as times of construction for the Abilene system and Colossus 1-2 demonstrate.

Thus compute planned for a year from now can get 6x cheaper if delayed now, and compute planned in 2+ years can get maybe 25x cheaper if delayed now. Planning for $300bn per year of compute in 2028 (25 GW, $1,250bn in capex) might mean you have $30bn per year of compute (2.5 GW) in 2026, that you plan for $70bn per year of additional compute to go online in 2027 (bringing the total to $100bn per year), and that you plan for $200bn per year of additional compute to go online in 2028. But if 2026 reveals that everything must stop, the 5.5 GW of additional planned compute for 2027 will only cost $11bn per year while delayed, and the 17 GW planned for 2028 might only cost $8bn per year.

So $300bn per year in planned compute (or $1,250bn in upfront capex) becomes $50bn per year in actual ongoing payments if compute has to be frozen at 2026 levels (with only 2.5 GW actually online, 60% of the cost going to live compute), the rest of the planned compute can be unfrozen as needed at the drop of a hat. And 50-70% gross margins for inference should be sufficient to pay for that, even if it might leave less than usual for research and training. In reality, it's very unlikely that there will need to be a complete freeze in buildout, so only a fraction of the delay overhead can plausibly materialize, making it correspondingly even less of an issue.