Necessary and Proportionate
In Which Civil Society is Caught Between a Cop and a Spy
A year ago and in much simpler times, Google was an uncomplicated ally in the fight against government surveillance, their transparency reports a shining example to the industry. At the behest of British NGO Privacy International, they spared no expense to fly fifty people to their offices in Brussels from around the world. Privacy International had decided it was high time the Internet freedom community stopped being reactive and started to make demands of their own, and from this, a coalition to make those demands was born.
I was in the room in Brussels. The talks were split between folks working in policy and technology, which may have been a partial source of the problem. The principles that come from that work include simple, sensible things like legality, necessity, proportionality, due process, transparency, and probable cause. They’re intended to constrain state surveillance to cases where it’s directed at catching a specific bad actor. They also include the notion that surveillance should be treated the same way everywhere—that just because someone is on the wrong side of a border shouldn’t mean they have fewer rights, and that states shouldn’t collude to arbitrage data with allies, each spying on the other's population and then sharing that data in an effort to gain surveillance data on their own people while skirting laws against domestic surveillance.
I’d be delighted to see these rules in place. They’d go a very long way toward stopping the abuses of the innocent via surveillance that we’ve seen law enforcement propagating. Even in today’s climate, seeing their adoption through for law enforcement will be a very hard battle. Institutions do not give up power willingly, even when the full force of public opinion falls against them. Once a technical capability has been fielded and proven useful, it is rare to see it revoked. This is the right and proper role of civil society, however — to fight that fight and see it through to the end. Ideally, the instruments of international law might come into play, and we would see a larger global consensus on the just powers of law enforcement.
But we won’t.
Once a technical capability has been fielded and proven useful, it is rare to see it revoked. This is the right and proper role of civil society, however — to fight that fight and see it through to the end.
We won’t, because law enforcement surveillance is inseparable from state intelligence, and because civil society has misunderstood both the dual nature of surveillance and the size of the demand they’re making.
The role of surveillance in law enforcement is to gather evidence required for a conviction. Enforcement of law requires the state to either know of or suspect a violation of the law has occurred. State intelligence has no such requirement; instead, it operates in the territory of unknown unknowns, attempting to parse the world in whatever ways the decisions of the state require.
Intelligence is a fundamental requirement for the modern state. Without it, the state is blind, unable to understand its place in an inherently adversarial structure. The state, which must attempt to preserve its territorial integrity, territorial monopoly on (some) force, and sovereignty of action, requires information about the actions of those adversaries known and unknown who may attempt to infringe on any of these structures.
Specifically, intelligence has a mission to both find previously unknown threats and to find secret information about the lawful activities of entities the state has a structurally adverse relationship with. For instance, within the boundaries set down by international treaty, China breaks no laws when it builds nuclear weapons; however, it is still not only in US interests but an existential requirement for the US to know as much as possible about those actions. Nothing that can impact the military and otherwise existential decisions of a state can be out of scope for the intelligence mission, because its mission is to support the state with the knowledge required for those decisions.
In a more perfect world, international law and treaty would guarantee this space of action, but the dream of rule of law at the geopolitical level died with Oppenheimer. The nuclear weapon represents an existential threat to states in a way that confirms that power not policy, will determine the outcome of international maneuvering. Since then, we’ve seen the cracks appear in the fictive structures built up to materialize its pretense—who still gives credence to the role of the International Criminal Court or the UN Security Council as we once did to the dream of the League of Nations?
This need for intelligence is implicit but unstated in the Westphalian compromise. The Peace of Westphalia, brought about in 1648, defined the agency of the state and a theory of non-interference. In practice, however, the Westphalian state set the rules around the power monopolies that the state must attempt to maintain. If a state cannot fulfill its Westphalian duties, it ceases to perform statedom; we call it a failed state if it is still independent and see it annexed into another state if it is not.
Modern telecommunications technology implied the development of modern telecommunications surveillance, because it moved the scope of action from the physical world (where intelligence, generally seen as part of the military mission, had acted) to the virtual world—including the scope of those actions that could threaten state power. While the public line may have been, as US Secretary of State Henry Stimson said in 1929, “gentlemen do not open each other’s mail”, you can bet that they always did keep a keen eye on the comings and goings of each other’s shipping traffic.
The real reason that surveillance in the context of state intelligence was limited until recently was because it was too expensive, and it was too expensive for everyone. The Westphalian compromise demands equality of agency as tied to territory. As soon as one side gains a significant advantage, the structure of sovereignty itself is threatened at a conceptual level — hence Oppenheimer as the death of any hope of international rule of law. Once surveillance became cheap enough, all states were (and will increasingly be) forced to attempt it at scale, as a reaction to this pernicious efficiency. The US may be ahead of the game now, but Moore’s law and productization will work their magic here.
The real reason that surveillance in the context of state intelligence was limited until recently was because it was too expensive, and it was too expensive for everyone.
Asking states to redraw the bright line between police surveillance and intelligence might be doable, but then we’re in effect asking for two separate systems of surveillance, producing twice the structural vulnerabilities in communication infrastructure. When you introduce a surveillance point in an otherwise secure system, you introduce a place where anyone, good or bad, may be able to compromise the system, as we saw when the Chinese targeted Google’s surveillance compliance systems. Not only that, but if the information exists somewhere, we’ll see a corrosive effect on both the barriers designed to keep intelligence and law enforcement apart and on intelligence community respect for rule of law as anything other than a chess piece in general. In the US we’ve seen this with the parallel construction revelations that have come to light, wherein NSA has passed intelligence product to the Drug Enforcement Agency’s Special Operations Division. The SOD passes this information on to officers in a number of US federal law enforcement agencies, who use it to ensure that they “get lucky” in their searches and thus construct an entirely parallel and “legal” chain of evidence.
In asking all states to confine themselves to only surveil as a law enforcement tactic, and to in effect do no international intelligence work (for intelligence can clearly not operate within these bound), the International Principles on the Application of Human Rights to Communication Surveillance ask for nothing less than the end of the Westphalian compromise and the creation of a new fundamental theory of geopolitical power and the monopoly on violence.
I support this whole-heartedly.
I’m not sure if my friends in civil society intended to bring on this renegotiation when they signed on to these principles. However, we dismantle the state without understanding the structures we propose to replace it at our extreme peril, and the principles do not speak to this topic at all.
If we wish to have a debate as to the basic structure of power we’d see replace the Westphalian state (and, necessarily also the structure that replaces the capitalist entity, for the two cannot be replaced separately), by all means, let’s do so. However, if we pretend to the utopian notion that we can decline this debate while still proposing the end to surveillance and thus intelligence, we reveal ourselves to not understand the first thing about the world in which we’re playing. As civil society, we do ourselves no favors by demonstrating our irrelevance.