The people who are not the target
Lawful interception touches people who are not its subject. Pretending otherwise would make everything else on this site harder to believe.
Two things are true at once.
Lawful interception is how serious crime is investigated in most democracies, and it works. It is also a capability that, misapplied, reaches into the lives of people who have done nothing and will never know they were touched. Both of those are true, and a vendor that only says the first is not worth trusting on the second.
Our position is that the safeguards have to live in the product. A policy document constrains an organization that already wanted to be constrained. Scope enforcement, automatic expiry, attribution and minimization constrain it on the day somebody decides the rules are inconvenient, which is the only day that matters.
What actually bounds the capability
Each of these is checkable in a deployment. Where a safeguard depends on the customer’s own authority rather than our engineering, it says so, because a vendor claiming credit for someone else’s process is the failure mode this page exists to avoid.
Collection is scoped, then enforced
The terms of an authorization become machine-enforced limits before collection starts. The system does not rely on an operator restraining themselves.
Expiry is automatic
When an authorization ends, collection ends. There is no step for anyone to forget, and no window where collection quietly outlives its warrant.
Minimisation is a workflow, not a policy
Material outside the scope of an order is handled through defined minimization rather than left in the record for someone to notice later.
Every access is attributable
Per-user attribution and a recorded reason for access on every action, in an append-only log. This is what makes an internal abuse investigation possible at all.
Oversight can see without operating
Inspectors and auditors can be granted genuine visibility into what was done, without acquiring the ability to task or collect.
The authorising process stays yours
We do not issue warrants, approve targets or run operations. The institution with the legal mandate does that, and we build so that its decisions are the ones the system obeys.
What we will not do
A commitment is only worth stating if it could cost something. Each of these has.
- Supply restricted capability to an institution without the legal authority and oversight to operate it.
- Build a covert channel that would let us observe a lawful customer’s operations.
- Weaken audit, attribution or expiry because a customer finds them inconvenient.
- Provide capability where the review reaches deferred or declined, on any commercial argument.
- Treat “the customer is responsible” as a complete answer to a misuse question.
Where our authority ends.
We decide whom we serve, what we supply and support, which boundaries apply, and when our own participation must stop. Inside the boundary that was agreed, an authorized institution governs its own operations. We do not run them and we do not covertly constrain them, because a vendor quietly steering a sovereign customer’s intelligence work is its own kind of accountability failure.
That is a real limit on what we can promise. What we can promise is that the capability we supply enforces the authority it was given, records what was done with it, and can be withdrawn by us if it is misused.
Raising a concern.
Anyone can raise a concern about how an S32 Technologies capability is being used, including people with no relationship to us and no connection to the customer. It reaches a human. Where a concern involves the conduct of a deployment, it is reviewed by someone outside the team that owns the account.
We do not run an external ethics board. Advisory bodies that can be overruled have a poor record in this industry, and we hold the decision, and the accountability for it, where it sits.