Let's play a mental game: How would you design a strategy for a battle you knew was already lost? What would the goal of the exercise be? What, to be blunt, would be the point?
This is not such an academic exercise. A recently announced attack has compromised the personal information of, ohhh, everyone. According to researchers at Upguard, a breach at Alteryx (a data analysis firm) has exposed personal information for roughly 120 million US households. Do a bit of math and it turns out that anyone old enough to have personally identifiable information has had it exposed.
Some of the news in this breach is immediately upsetting. One of Alteryx's partners is Experian, one of the Big 3 credit reporting firms, so it can be assumed that there is at least a path from Alteryx to a lot of credit card and other very sensitive information. This economic information is an immediate concern and a massive pain in the patoot to deal with.
Much of the data bulk in the release, though, was US Census data. While much of that is publicly available, its existence at Alteryx demonstrates why it represents a "force multiplier" for the other data stolen in the attack.
In the big data world, the kind of data collected in a US Census can be combined with sensitive PII, social media information and data from other government databases to create very rich, complete data portraits of individuals. Those portraits can then be used for purposes that range from identity theft to highly successful spear phishing campaigns.
But all of that begs the original question: If the data you're protecting has already been stolen from someone else, do you still need to protect it? I think the answer is rather more nuanced than a simple "yes/no" response can convey.
The nuance comes because news of massive data breaches reinforces the idea that data should be prioritized when it comes to security. Not everything is sensitive to the same degree. Names and addresses are one thing, information on transactions a level up (especially if those transactions are, in themselves, sensitive), then comes financial information and finally, at the top of the pyramid, is the Social Security (or other government system) number.
Cybersecurity professionals have long pushed the notion that their organizations should prioritize data for security purposes and the critical nature of that advice just keeps climbing. But it's also important to admit that even the most publicly available data should be protected when it's in your hands.
The reason gets back to the big data connection. The fact that an individual is a customer of your organization adds to the richness of the data set available on that person. While it might not be, in and of itself, enough detail to allow identity theft, it can certainly amplify the impact of other data in providing a fake identity to someone.
Just as important, this news is a great reason to ask whether data your organization is holding should be kept at all. Prioritizing data sensitivity should contain a step in which some data is not stored at all or is aged out of the system very, very quickly. We've become data hoarders in our businesses, certain that there's gold in the data if we just hold it long enough and figure out new ways to sift away the chaff. What we're finding is that the gold may be illusory, but the cost of protecting it is very real indeed.