Lately I’ve been writing about the policy issues surrounding government wiretapping programs that algorithmically analyze large amounts of communication data to identify messages to be shown to human analysts. (Past posts in the series: 1; 2; 3; 4; 5; 6; 7.) One of the most frequent arguments against such programs is that there will be too many false positives – too many innocent conversations misidentified as suspicious.

Suppose we have an algorithm that looks at a set of intercepted messages and classifies each message as either suspicious or innocuous. Let’s assume that every message has a true state that is either criminal (i.e., actually part of a criminal or terrorist conspiracy) or innocent. The problem is that the true state is not known. A perfect, but unattainable, classifier would label a message as suspicious if and only if it was criminal. In practice a classifier will make false positive errors (mistakenly classifying an innocent message as suspicious) and false negative errors (mistakenly classifying a criminal message as innocuous).

To illustrate the false positive problem, let’s do an example. Suppose we intercept a million messages, of which ten are criminal. And suppose that the classifier correctly labels 99.9% of the innocent messages. This means that 1000 innocent messages (0.1% of one million) will be misclassified as suspicious. All told, there will be 1010 suspicious messages, of which only ten – about 1% – will actually be criminal. The vast majority of messages labeled as suspicious will actually be innocent. And if the classifier is less accurate on innocent messages, the imbalance will be even more extreme.

This argument has some power, but I don’t think it’s fatal to the idea of algorithmically classifying intercepts. I say this for three reasons.

First, even if the majority of labeled-as-suspicous messages are innocent, this doesn’t necessarily mean that listening to those messages is unjustified. Letting the police listen to, say, ten innocent conversations is a good tradeoff if the eleventh conversation is a criminal one whose interception can stop a serious crime. (I’m assuming that the ten innocent conversations are chosen by some known, well-intentioned algorithmic process, rather than being chosen by potentially corrupt government agents.) This only goes so far, of course – if there are too many innocent conversations or the crime is not very serious, then this type of wiretapping will not be justified. My point is merely that it’s not enough to argue that most of the labeled-as-suspcious messages will be innocent.

Second, we can learn by experience what the false positive rate is. By monitoring the operation of the system, we can see learn how many messages are labeled as suspicious and how many of those are actually innocent. If there is a warrant for the wiretapping (as I have argued there should be), the warrant can require this sort of monitoring, and can require the wiretapping to be stopped or narrowed if the false positive rate is too high.

Third, classification algorithms have (or can be made to have) an adjustable sensitivity setting. Think of it as a control knob that can be moved continuously between two extremes, where one extreme is labeled “avoid false positives” and the other is labeled “avoid false negatives”. Adjusting the knob trades off one kind of error for the other.

We can always make the false positive rate as low as we like, by turning the knob far enough toward “avoid false positives”. Doing this has a price, because turning the knob in that direction also increases the number of false negatives, that is, it causes some criminal messages to be missed. If we turn the knob all the way to the “avoid false positives” end, then there will be no false positives at all, but there might be many false negatives. Indeed, we might find that when the knob is turned to that end, all messages, whether criminal or not, are classified as innocuous.

So the question is not whether we can reduce false positives – we know we can do that – but whether there is anywhere we can set the knob that gives us an acceptably low false positive rate yet still manages to flag some messages that are criminal.

Whether there is an acceptable setting depends on the details of the classification algorithm. If you forced me to guess, I’d say that for algorithms based on today’s voice recognition or speech transcription technology, there probably isn’t an acceptable setting – to catch any appreciable number of criminal conversations, we’d have to accept huge numbers of false positives. But I’m not certain of that result, and it could change as the algorithms get better.

The most important thing to say about this is that it’s an empirical question, which means that it’s possible to gather evidence to learn whether a particular algorithm offers an acceptable tradeoff. For example, if we had a candidate classification algorithm, we could run it on a large number of real-world messages and, without recording any of those messages, simply count how many messages the algorithm would have labeled as suspicious. If that number were huge, we would know we had a false positive problem. We could do this for different settings of the knob, to see where we had to get an acceptable false positive rate. Then we could apply the algorithm with that knob setting to a predetermined set of known-to-be-criminal messages, to see how many it flagged.

If governments are using algorithmic classifiers – and the U.S. government may be doing so – then they can do these types of experiments. Perhaps they have. It doesn’t seem too much to ask for them to report on their false positive rates.

When the NSA’s wiretapping program first came to light, the White House said, mysteriously, that they didn’t get warrants for all of their wiretaps because doing so would have been impractical. Some people dismissed that as empty rhetoric. But for the rest of us, it was a useful hint about how the program worked, implying that the wiretapping was triggered by the characteristics of a call (or its contents) rather than following individuals who were specifically suspected of being terrorists.

As I wrote previously, content-based triggering is a relatively recent phenomenon, having become practical only with the arrival of the digital revolution. Our laws about search, seizure, and wiretapping mostly assume the pre-digital world, so they don’t do much to address the possibility of content-based triggering. The Fourth Amendment, for example, says that search warrants must “particularly describ[e] the place to be searched, and the persons or things to be seized.” Wiretapping statutes similarly assume wiretaps are aimed at identified individuals.

So when the NSA and the White House wanted to do searches with content-based triggering, there was no way to get a warrant that would allow them to do so. That left them with two choices: kill the program, or proceed without warrants. They chose the latter, and they now argue that warrants aren’t legally necessary. I don’t know whether their legal arguments hold water (legal experts are mostly skeptical) but I know it would be better if there were a statute that specifically addressed this situation.

The model, procedurally at least, would follow the Foreign Intelligence Surveillance Act (FISA). In FISA, Congress established criteria under which U.S. intelligence agencies could wiretap suspected spies and terrorists. FISA requires agencies to get warrants for such wiretaps, by applying to a special secret court, in a process designed to balance national security against personal privacy. There are also limited exceptions; for example, there is more leeway to wiretap in the first days of a war. Whether or not you like the balance point Congress chose in FISA, you’ll agree, I hope, that it’s good for the legislature to debate these tradeoffs, to establish a general policy, rather than leaving everything at the discretion of the executive branch.

If it took up this issue, Congress might decide to declare that content-based triggering is never acceptable. More likely, it would establish a set of rules and principles to govern wiretaps that use content-based triggering. Presumably, the new statute would establish a new kind of warrant, perhaps granted by the existing FISA court, and would say what justification needed to be submitted to the court, and what reporting needed to done after a warrant was granted. Making these choices wisely would mitigate some of the difficulties with content-based triggering.

Just as important, it would create a constructive replacement for the arguments over the legality of the current NSA program. Today, those arguments are often shouting matches between those who say the program is far outside the law, and those who say that the law is outdated and is blocking necessary and reasonable intelligence-gathering. A debate in Congress, and among citizens, can help to break this rhetorical stalemate, and can re-establish the checks and balances that keep government’s power vital but limited.