Last week California became the first state to enact a law specifically addressing phishing. Phishing, for those of you who have been away from the internet for the past few years, is when an attacker sends you an e-mail falsely claiming to be a legitimate business in order to trick you into giving away your account info — passwords, mostly. When this is done by hacking DNS, it's called pharming.
Financial companies have until now avoided taking on phishers in a serious way, because it's cheaper and simpler to pay the costs of fraud. That's unacceptable, however, because consumers who fall prey to these scams pay a price that goes beyond financial losses, in inconvenience, stress and, in some cases, blots on their credit reports that are hard to eradicate. As a result, lawmakers need to do more than create new punishments for wrongdoers — they need to create tough new incentives that will effectively force financial companies to change the status quo and improve the way they protect their customers’ assets. Unfortunately, the California law does nothing to address this.
The new legislation was enacted because phishing is a new crime. But the law won't help, because phishing is just a tactic. Criminals phish in order to get your passwords, so they can make fraudulent transactions in your name. The real crime is an ancient one: financial fraud.
These attacks prey on the gullibility of people. This distinguishes them from worms and viruses, which exploit vulnerabilities in computer code. In the past, I've called these attacks examples of “semantic attacks” because they exploit human meaning rather than computer logic. The victims are people who get e-mails and visit websites, and generally believe that these e-mails and websites are legitimate.
These attacks take advantage of the inherent unverifiability of the internet. Phishing and pharming are easy because authenticating businesses on the internet is hard. While it might be possible for a criminal to build a fake bricks-and-mortar bank in order to scam people out of their signatures and bank details, it's much easier for the same criminal to build a fake website or send a fake e-mail. And while it might be technically possible to build a security infrastructure to verify both websites and e-mail, both the cost and user unfriendliness means that it'd only be a solution for the geekiest of internet users.
These attacks also leverage the inherent scalability of computer systems. Scamming someone in person takes work. With e-mail, you can try to scam millions of people per hour. And a one-in-a-million success rate might be good enough for a viable criminal enterprise.
In general, two internet trends affect all forms of identity theft. The widespread availability of personal information has made it easier for a thief to get his hands on it. At the same time, the rise of electronic authentication and online transactions — you don't have to walk into a bank, or even use a bank card, in order to withdraw money now — has made that personal information much more valuable.
The problem of phishing cannot be solved solely by focusing on the first trend: the availability of personal information. Criminals are clever people, and if you defend against a particular tactic such as phishing, they'll find another. In the space of just a few years, we've seen phishing attacks get more sophisticated. The newest variant, called “spear phishing,” involves individually targeted and personalized e-mail messages that are even harder to detect. And there are other sorts of electronic fraud that aren't technically phishing.
The actual problem to be solved is that of fraudulent transactions. Financial institutions make it too easy for a criminal to commit fraudulent transactions, and too difficult for the victims to clear their names. The institutions make a lot of money because it's easy to make a transaction, open an account, get a credit card and so on. For years I've written about how economic considerations affect security problems. They can put security countermeasures in place to prevent fraud, detect it quickly and allow victims to clear themselves. But all of that's expensive. And it's not worth it to them.
It's not that financial institutions suffer no losses. Because of something called Regulation E, they already pay most of the direct costs of identity theft. But the costs in time, stress and hassle are entirely borne by the victims. And in one in four cases, the victims have not been able to completely restore their good name.
In economics, this is known as an externality: It's an effect of a business decision that is not borne by the person or organization making the decision. Financial institutions have no incentive to reduce those costs of identity theft because they don't bear them.
Push the responsibility — all of it — for identity theft onto the financial institutions, and phishing will go away. This fraud will go away not because people will suddenly get smart and quit responding to phishing e-mails, because California has new criminal penalties for phishing, or because ISPs will recognize and delete the e-mails. It will go away because the information a criminal can get from a phishing attack won't be enough for him to commit fraud — because the companies won't stand for all those losses.
If there's one general precept of security policy that is universally true, it is that security works best when the entity that is in the best position to mitigate the risk is responsible for that risk. Making financial institutions responsible for losses due to phishing and identity theft is the only way to deal with the problem. And not just the direct financial losses — they need to make it less painful to resolve identity theft issues, enabling people to truly clear their names and credit histories. Money to reimburse losses is cheap compared with the expense of redesigning their systems, but anything less won't work.
Bruce is right. Let me put it this way. Sites must move as quickly as they can towards what Toby Stevens calls “Data Rejection“, minimizing retention of individually identifying information. They must ensure that PII which needs to be retained is encrypted, decipherable only through systems which are quaranteened from the Internet and have the proper operational controls.
The InfoCard system has been devised to allow companies to practice Data Rejection. It uses cryptography to recognize digital relationships so personal identifying information can be made available to an internet site while a transaction is in progress but not be stored there – except, perhaps, in encrypted audit logs.