A government is on the cap table now
CuspAI raised $450M with the UK Sovereign AI Venture Fund alongside Kleiner Perkins and Bezos Expeditions. State money is a normal co-investor. Price what it asks for.
Latest
-
-
Someone raised $75M to question the Fortune 500 spending
Freehand put AI agents in charge of supply-chain spend at Meta, Pfizer and Unilever. The same audit takes an afternoon on your ledger, and diligence will do it anyway.
-
The 80% gross margin is gone, and diligence already knows
AI-native products run near 52% gross margin against 75% to 85% for classic SaaS. Bring the COGS breakdown yourself, before someone asks for it.
-
The record year is not the market you are raising into
H1 2026 set a $510B record and two companies took nearly half of it. Almost none of that is a comparable for your round. Here is how to build the set that is.
-
They underwrite retention, not growth
The 2026 Series A bar for AI companies runs near $3.5M ARR and 120% net revenue retention. Growth gets you the meeting. Retention gets the money.
-
When your investor is also your vendor
Nvidia has announced over $540B of deals in 2026 and takes stakes in the customers who buy from it. That structure has reached your seed round. Price it before you sign.
-
Your deck is read by a machine before a partner sees it
85% of dealmakers now run AI on daily work and 82% use it for sourcing. Your deck gets summarised and scored before a human opens it. Write for that pass.
-
Your investor update is a data feed now
Funds run agents over board decks and founder updates to pull KPIs automatically. Write updates that parse cleanly, because inconsistency across six months gets caught.
-
A list of investors is not an asset
Clay is worth billions for joining data sources together. The same logic says your 300-name investor list is close to worthless, and what to do instead.
-
The bottleneck was never the model. It was permission.
Arcade raised $60M so AI agents can prove who did what. Your raise has the same missing record, and a spreadsheet fixes it.
-
The money moved to the long, boring tasks
Sail Research raised $80M for agents that run for days, not seconds. The same shift says you are automating the wrong half of your raise.
-
The other side of your term sheet reads faster than you do
Harvey hit an $11B valuation selling legal agents to firms. Here is the prompt that makes your hour with a lawyer cheaper.
-
Someone raised $28M to break AI agents on purpose
Coval simulates the calls that make a voice agent fail. The same idea works on a raise, and the prompt is free.
-
AI came for the reference call, not the deck
Diligence automation went after the half of the raise founders do not control. What that changes about how you prepare.
-
Why this blog exists
The short version goes on LinkedIn. The whole argument goes here.