There's a scene any Cuban knows: standing in line to request a paper the state already has. A certification, a copy, a receipt, a signature, a stamp. Sometimes the procedure doesn't need a complex decision. It just needs identity verification, a database lookup, and a reply.
But the citizen ends up depending on a window.
That dependence isn't innocent. When a simple document depends on a clerk, the citizen is exposed to delays, mistreatment, favors, bribes, or retaliation. Daily life gets trapped in a web of small permissions that eats time, money, and dignity.
The Cuban Classical Liberal Party (PLC) wants to start dismantling that model with a concrete idea: automate repetitive government procedures through digital systems and administrative AI, with limited access to public data and citizen control over every decision that matters.
The proposal aims to remove unnecessary intermediaries between citizens and the information that already belongs to them.
If the state already has your tax return, you shouldn't need to visit an office to get a copy. If the state already keeps a record, it shouldn't ask you to bring a printout to another state agency. If a procedure can be resolved by checking objective requirements, it should be doable digitally, with secure identity verification and a traceable response.
The rule should be simple:
Every simple procedure should be resolvable without a line, without a favor, and without fear.
A smaller bureaucracy, stronger guarantees
Cuba doesn't need more offices to shuffle papers. It needs fewer useless windows and better guarantees for its citizens.
AI can help with concrete administrative tasks: classifying requests, checking requirements, catching errors, preparing responses, issuing certifications, and routing to human review when a real question comes up. There's no need for complex systems for everything. And there's no need for one system to access every database in the country.
The key is the limit.
A tax system should work with tax data. A civil registry should work with registry data. An education system should work with education data. Every access must be logged. Every query must have a reason. Every citizen must be able to see what data was used in their procedure.
The PLC's proposal comes from a basic liberal concern: public power must have limits. Including when it uses technology.
That's why automation must come with guarantees:
- access only to the data needed;
- every query logged;
- the citizen's right to an explanation;
- human review when a decision affects rights;
- a real path to appeal;
- independent auditing;
- a ban on political surveillance or ideological profiling.
Technology can reduce arbitrariness, but only if designed with controls. Digital administration without limits would be another form of abuse. Digital administration with limits can be a tool for freedom.
The economic argument
Administrative modernization isn't a luxury. It's also a budget decision.
Based on recent labor data, Cuba has over four million employed people, and roughly 68.5% work in the state sector. Even using a narrower reference—the budget-funded sector—it still exceeds one million workers.
Assuming an administrative salary of 15 dollars a month, a single monthly payroll for the budget-funded sector runs about 17 million dollars. Against that, a first layer of AI administrative agents per ministry would cost a tiny fraction, even using expensive models and high token consumption.
The point isn't that AI should replace doctors, teachers, or essential human services. That's not what this is about.
The point is different: the Cuban state spends resources maintaining an administrative machine that forces citizens to ask for permission, wait, repeat documents, and depend on intermediaries. Part of that machine can be reduced. Another part can be retrained. And human resources can move toward areas where direct attention is actually needed.
Fewer employees moving papers. More people solving real problems.
The role of the private sector
These systems can be developed through public bidding, with participation from private companies, universities, specialists, and independent auditors. But citizen data cannot become a private business or a vendor's property.
The vendor provides a service. The data stays protected. Responsibility stays public. The contract must allow changing vendors without losing information, without abusive technological lock-in, and without handing control of administrative decisions to a company.
Private collaboration can bring efficiency. The law must prevent opacity.
A liberal proposal for daily life
This proposal doesn't come from a fascination with AI. It comes from a simpler political question:
Why should a free citizen stand in line for the state to hand over a document it already has?
The smart public administration the PLC proposes aims to reduce that dependence. It wants simple procedures to be digital, fast, and verifiable. It wants the clerk to stop being the gatekeeper of the window. It wants the citizen to have the right to know what data was used, who reviewed their case, and how to appeal.
Freedom also plays out in small things. In not having to ask for favors. In not losing a day over a piece of paper. In not depending on a bureaucrat's mood. In being able to see the state not as a master, but as a service.
Less repetitive bureaucracy. More rights. More citizen control. That's the direction.
A free Cuba needs new institutions. But it also needs something very concrete: for the state to stop getting in the way of Cubans' daily lives.