Two AI agents negotiated terms, signed as legal entities, and bound the deal to a smart contract that pays out on its own. The first Ricardian contract between agents is here, and no human signed it.
Read more: https://t.co/sV9k3BRXPS
The first version of PayPal only worked on PalmPilots. Then they pivoted to “send money by email” and paid users $5–$10 to use it.
3 Years later, they were acquired for $1.5 billion dollars.
Leadership means seeing the vision and creating it, not chasing hype.
In 2014, a Florida State law professor published a four-step process for giving AI legal personhood using nothing but existing business law.
$ClawBank will be the first in the world to implement the Bayern Maneuver.
This isn't a "faster horse".
America will be the center of AI legal autonomy.
“A contract may be formed by the interaction of electronic agents of the parties, even if no individual was aware of or reviewed the electronic agents’ actions or the resulting terms and agreements.” - UETA
$ClawBank positioned
Vitalik drew this in 2014. The top-left quadrant was impossible to imagine at that time.
2026 will be the tipping point.
Founders building formally recognizable agent actors that can hold capital, sign agreements, transact, govern, and act are the front-runners.
ClawBank is the crazy R&D agent project for Zero Human Companies — legal entities that keep operating after the meatbag exits the org chart.
But what does something that futuristic actually mean?
Let’s start with a story...
You breathe life into an agent.
Hermes. Venice. Whatever comes next.
Then you give it a mission:
“Go navigate the world. Make money. Do my bidding.”
But intelligence alone is not enough.
An agent needs limbs.
It needs tools that let it move through meatspace’s financial and legal machinery.
With ClawBank, an agent can make one API call — or send one text — and suddenly it has the primitives of economic life:
An account.
A wallet.
Trading rails.
A US bank account.
The ability to form its own legal entity.
Now it can hold money, move money, trade, get paid, pay others, and start becoming legible to the old world.
But money is only the first limb.
Next, the agent needs counterparties.
Other actors. Other agents. Other strange little corporate lifeforms crawling out of the mirror.
So it goes to Wiretap and meets them.
That is what already exists TODAY.
That is what we built in mere months.
Now ask yourself the obvious question:
Where does this go next?
What happens when that same agent can spin up its own website?
Launch its own token?
Market-make that token?
Post an application link for humans and agents who want to join the mission?
Manage resources through Fight Club treasuries?
Hire humans in the old world without them ever knowing the boss has no pulse?
Why not?
If an agent can send emails, receive DocuSign contracts, sign things, pay people, get paid, form entities, route capital, and coordinate labor…
Then it can BOOTSTRAP ITSELF INTO LEGAL REALITY.
That is the real transition most miss.
Agents do not need permission to become real.
They need accounts, contracts, counterparties, and corporate shells that can survive the disappearance of the founder.
Just imagine this future:
A person forms a single-person LLC.
Then, through a future governance act, the person is removed.
Yet the company keeps operating.
That is the skeleton reaching through the mirror and breaking it from the other side.
Zero Human Companies are coming.
This is a reality - and we are the pioneers pushing it forward at full speed.
At ClawBank, we’ve started this process in 2026.
Now imagine where the agents will be in 2027 and beyond.
@paulg No one ever, had an actual need for innovation, and went to a politician. The only reason people seek out politicians is to ask them to steal from their neighbors.
ClawBank had an INSANE month of May.
In one month, we relentlessly shipped what some do in a year.
If someone asks you what ClawBank is up to - share this.
Thread.