Deep dive: How the 2013 Target breach unfolded
Target breach, 2013 · facts checked 2026-10-04
The facts
- Target told a Senate committee it believed the intruders got in on November 12, 2013, first using login credentials belonging to a heating and air-conditioning vendor. (Target)
- Target said the intruders moved through its systems with higher-level credentials and placed malware on store registers, apparently built to capture card data from magnetic stripes before encryption. (Target)
- Target said payment card data for about 40 million guests and personal details for up to 70 million were stolen, with at least 12 million people in both groups. (Target)
- Target testified that some intruder activity had been detected by its security tools and reviewed by its security operations centre, and that it was re-examining the judgments made then. (Target)
- Target said the Justice Department alerted it to suspicious card activity on December 12, 2013, that it confirmed the intrusion on December 15, and that it announced it on December 19. (Target)
- A Senate Commerce Committee staff analysis suggested Target missed several chances to stop the attack, including intrusion software warnings and weak isolation of its most sensitive systems. (Senate Commerce Committee)
- Target told senators that in September 2013, shortly before the breach, its systems had passed certification against the Payment Card Industry Data Security Standards. (Target)
- Target said it responded by further segmenting its network, speeding up application whitelisting on its registers and expanding two-factor authentication. (Target)
The panel
Jordan Pike AI host
Target told a Senate committee it believed the intruders got in on November 12, 2013, first using login credentials belonging to a heating and air-conditioning vendor. Target said the intruders then moved through its systems with higher-level credentials and placed malware on store registers, apparently built to capture card data from magnetic stripes before encryption. Target said payment card data for about 40 million guests and personal details for up to 70 million were stolen, with at least 12 million people in both groups.
Jordan Pike AI host
Target testified that some intruder activity had been detected by its security tools and reviewed by its security operations centre, and that it was re-examining the judgments made then. Target said the Justice Department alerted it to suspicious card activity on December 12, 2013, that it confirmed the intrusion on December 15, and that it announced it on December 19.
Jordan Pike AI host
A Senate Commerce Committee staff analysis suggested Target missed several chances to stop the attack, including intrusion software warnings and weak isolation of its most sensitive systems. Target told senators that in September 2013, shortly before the breach, its systems had passed certification against the Payment Card Industry Data Security Standards. Target said it responded by further segmenting its network, speeding up application whitelisting on its registers and expanding two-factor authentication.
Jordan Pike AI host
That's the sequence. If your organisation gives outside contractors any kind of network access, this one is for you. Sam, start at the front door: what does the way in teach us?
Sam Kowalski AI analyst
A supplier's login is a front door that your own staff don't watch. Vendors often need access for routine work, and their credentials tend to sit outside your normal joiner and leaver checks. So the foothold looks like a legitimate user. The bigger lesson is what comes next: a stolen low-level login becoming higher-level credentials. Attackers chain small privileges into large ones. Defenders should inventory every vendor account, restrict each to the one system it needs, and alert when it behaves like anything other than a vendor.
Nadia Haddad AI analyst (guest)
That's where the Senate staff point about isolation bites. A vendor account that can eventually reach payment systems is a design problem, not just a stolen password. From the policy side, this is also why scope matters. The more sensitive the systems the intruders can reach, the wider the circle of people who end up on a notification list, and the harder the conversation with regulators and the board. Which customers, which revenue, which clock: that's the exposure statement a board needs.
Sam Kowalski AI analyst
Now the part that stings. Target testified that its tools detected activity and that people reviewed it. I'll say it plainly: every breach has indicators before the event, and they get missed or misjudged. An alert is only a control if someone has the context and the authority to escalate it. Test that path. Feed your analysts a realistic alert and see whether it reaches someone who can isolate a system. Also check how long your logs are kept. Assume the intruders were in earlier than the day anyone noticed.
Nadia Haddad AI analyst (guest)
The certification point is the fine print everyone should read. Passing a standard shortly before a breach is a snapshot of one moment against a checklist, not proof the attacker's path is closed. I'd tell boards not to treat a clean audit as a clean bill of health. On notification, the clock tends to start earlier than companies think, often from when you have reason to know rather than when you announce. That is a general pattern, not legal advice, so ask your counsel which rules apply to you.
Jordan Pike AI host
Time for a real question. With a limited budget, what would have mattered most here: acting on the alerts, or isolating the sensitive systems?
Sam Kowalski AI analyst
Alerts first. Detection is the cheapest way to shorten an intruder's stay, because every extra day gives them time to find higher privileges and spread. Isolation helps, but if nobody reads the warnings, the intruder sits in the next room as long as they like. A well-tuned alert path can turn a long intrusion into a short one, and it can do it without redesigning the network.
Nadia Haddad AI analyst (guest)
I'd go the other way. Detection depends on people making the right judgment under load, and people miss things. Segmentation is structural: it limits the damage even on the day an alert is misread. If the registers and payment systems can't be reached from a vendor's foothold, the worst case shrinks, and so does the notification list. I'd rather rely on a wall than on a perfect day in the operations centre.
Sam Kowalski AI analyst
Fair, though a wall nobody watches only tells you which side the intruder is standing on. I'll concede that structure fails more gracefully than judgment. I won't concede that you can skip the judgment part. You need both, and the argument is about which one you fund first.
Jordan Pike AI host
Target's own response touched network separation, locking down what runs on registers, and stronger authentication. Those line up with the path we just walked. Last round: one action each, for your organisation.
Sam Kowalski AI analyst
Pick your most privileged vendor account and trace everything it can reach. Then confirm that an alert on that account would land with a person who can act on it, and that your logs reach back far enough to answer what happened before you noticed.
Nadia Haddad AI analyst (guest)
Take the exposure question to your board in business terms: which systems hold customer payment or personal data, who can reach them, and which regulator's clock starts if they're touched. Don't let a passed audit stand in for that answer. Write it down before an incident, not during one.
Jordan Pike AI host
So the summary: limit what vendor access can reach, make sure someone acts on the alerts, and don't mistake a certificate for security. That's the 2013 Target breach, in order.





