The incident involved theft of more than 100 million customer records, 140,000 social security numbers and 80,000 linked bank details of Capital One customers, allegedly stolen by a single, experienced engineer, according to court filings in Seattle.
The details set it apart from breaches of companies like Equifax and Marriott, which were attacked from the outside by criminals with a nation-state connection. It’s also different than the spate of ransomware attacks against major U.S. cities, which were likely committed by groups of individuals outside the U.S.
Instead, according to the indictment of Paige Thompson, she was able to exploit a loophole in a Capital One cloud server’s firewall to gain access to the information.
An Amazon spokesperson confirmed Thompson worked for Amazon, but left in 2016. The breach took place between March and July this year. Capital One confirmed in a statement Monday that the incident was related to a misconfigured application firewall and not an issue with cloud infrastructure.
”[Amazon Web Services] was not compromised in any way and functioned as designed,” Amazon said in a statement, adding that the reason for the breach was a misconfiguration of firewall settings on a web application, managed on the cloud server by Capital One, not a vulnerability in the cloud server itself.
The incident, which is still unraveling, will bring up major issues facing the biggest tech companies, cloud firms and banks, namely how to control who has access to sensitive consumer data and detect insiders who may go rogue.
An unlikely scenario
In many ways, it’s the nightmare scenario for a large company. Banks like Capital One have in recent years become much more adept at protecting against outside threats that target sensitive personal data. But protecting against a single individual bent on destruction and with even a modicum of access can be much harder.
According to the indictment, Thompson exploited a misconfigured firewall in a cloud server used by Capital One. She allegedly used a Tor browser, which anonymizes a person’s online activities, to gain this access. She also used a virtual private network known as IPredator to further obscure her activities, according to the indictment.
All of these factors combined with the possibility of insider knowledge means this incident will be closely watched by cybersecurity professionals and banks, particularly to see whether there was any way Capital One could have avoided the incident under the circumstances.
“Capital One had some good security practices in place ,” said Sam Curry, chief security officer of cybersecurity company Cybereason. “As a positive, they made an arrest quickly and there is a chance to minimize damage. Normally, it’s months, years or never in terms of arrests and accountability of the criminals. Finding things sooner in the life cycle always limits the impact and damage to the innocent.”
Capital One’s stock was down around 5% in early trading.