Order is not an end in itself. It is a means. When the means becomes more valuable than the end, the society we sought to protect is the first to disappear.
There are moments when technique ceases to be technique and becomes politics. The CSAR proposal has been under negotiation for four years, hundreds of pages of cryptographic analysis, dozens of legal opinions. And yet, what is rarely discussed with honesty is what is at stake: not whether scanning is possible, but whether it is desirable.
The investigations in this series have explored the promises and limits of each technical approach. FHE does not scale. ZKPs cannot detect what is not catalogued. CSS does not preserve encryption. External encryption does not survive a law that prohibits it. These are not political conclusions. They are mathematical conclusions. But from these conclusions follows a question that technique cannot answer: does it remain reasonable to insist on scanning when every technical path has proven incompatible with the privacy it claims to respect?
The core of the discord
Chat Control is not, at its core, a debate about technology. It is a debate about what kind of society we want to build. Technology is the vehicle, but the question is another: is it justified to sacrifice fundamental rights of the entire population for the promise of detecting a crime?
The European Commission's answer appears to be yes. And that answer deserves examination, not because the crime it seeks to prosecute is not atrocious —it is— but because the proposed remedy is structurally incompatible with the principle it claims to protect.
End-to-end encryption is not a negotiable feature. It is not an extra that can be disabled without consequences. It is the border between private life and surveilled life. And what the CSAR proposal asks, in essence, is that we accept that this border be porous —that the guard may look in whenever deemed necessary, with the promise that it will only look at what it should.
The imperceptible slope
One of the most uncomfortable lessons from the research in this series is the recurrence of a pattern: mass surveillance mechanisms tend to expand until they occupy all available legal space. Not by conspiracy, but by institutional engineering. A system designed to detect CSAM can, with a classifier update, detect terrorism. With another, hate speech. With another, criticism of the government. Each expansion is individually reasonable; the aggregate result is not.
The history of CALEA, of SWIFT, of every European internal security law, shows the same movement: a limited purpose, a progressive expansion, a final scope that the original drafters assured would never happen. The CSAR proposal asks us to trust that this time will be different. There are no historical reasons for that trust.
The unnecessary sacrifice
Alternatives to mass scanning exist. They are not perfect. They do not detect all CSAM. But they exist: reporting systems, hotlines, detection in public uploads, international cooperation, targeted investigation with judicial guarantees. None require the sacrifice of encryption.
The question that no Commission document answers satisfactorily is: why is the most intrusive measure precisely the one demanded when less harmful alternatives exist? The answer that emerges after reading the positions is uncomfortable: because the mechanism —a mandatory scanning infrastructure across all messaging platforms— has a scope that transcends any specific crime. CSAM is the justification. Control is the mechanism. And control, once installed, is not uninstalled.
CSAM as justification: "Only to protect the children"
Required mechanism: Scanning of all communications
Possible expansion: Any classifiable content
Real limit: None (except what the law imposes)
And who watches the law when it expands?
What is at stake
If Chat Control is approved in its current form, something is lost that cannot be recovered with an amendment: the certainty that private communications are truly private. Not because one has something to hide, but because a message that cannot be sent without being inspected is not a private message. It is a public statement with a guaranteed audience.
And that, perhaps, is the deepest transformation the regulation proposes: moving from a space where privacy is the rule to one where it is the exception that must be justified. The world that encryption protected was not the world of illegitimate secrets, but the world of one's own spaces. It was the possibility of saying something without anyone else hearing it, not because it is illegal, but because it is one's own.
A decision that is definitively not technical
In the end, cryptographers have already said what they had to say: encrypted content cannot be scanned without breaking encryption. That conclusion will not change with more research, better algorithms, or faster hardware. It is a structural limitation, not a technological one.
The decision that remains is political. One can accept that encryption means some communications will be inaccessible to authority, and seek other ways to protect minors. Or one can sacrifice encryption in the name of protection, and accept the consequences —known, documented, predictable— of living in a digital space where every message can be inspected.
The CSAR proposal seems to lean toward the latter. And that leaning, seen from the research that underpins this series, seems not only excessive but counterproductive: a measure that sacrifices fundamental rights for a promise of control that technology cannot deliver, and in doing so weakens both trust in institutions and the security it claims to protect.
Chat Control Series:
- Chat Control — Definition and context — The starting point: what CSAR proposes, its legislative journey, and the positions at stake.
- Encryption and surveillance — Why mass scanning breaks E2EE — Why mandatory scanning and end-to-end encryption cannot coexist.
- Homomorphic cryptography — Promises and limits — Why FHE is not the technical solution the Commission is looking for.
- Zero-knowledge proofs — Applications and limits — ZKP, PSI, and why they do not solve detection without breaking privacy.
- External encryption and shared keys — The user sovereignty approach and why it does not scale as a systemic solution.
- Alternatives to mass scanning — Technical, political and hybrid strategies to preserve privacy.
- The price of control — A perspective on sacrificing fundamental rights in the name of security.