Fake president’s fraud or social engineering fraud is occurring at an increasingly alarming rate. In one fell swoop, millions of dollars can be embezzled and transferred to offshore accounts.
The scheme is shockingly simple. A key employee in a victim's finance department receives an email from what appears to be the company’s CEO. The CEO is negotiating a confidential transaction involving the acquisition by the company and requests that the employee wire $1,000,000 to an overseas account to cover due diligence costs. The email also copies what appears to be legitimate local legal counsel. The CEO underlines the urgency and secrecy of the matter and does not want the employee to discuss the matter with anyone else. Although infrequent, such acquisitions are not unheard of at the company.
The employee, eager to please, and flattered by the responsibility bestowed upon him by the CEO, rushes to make the wire transfer to the offshore account. Happy to have helped, the employee later takes a closer look at the email. The CEO took an uncharacteristically formal tone and the email doesn't seem quite right anymore. The “i” in his email address appears to have been replaced by a “1”. Suddenly, reality sinks in: the company has just been defrauded and the money will likely never be recovered.
In this scenario, the insurance carrier is usually immediately advised and the only question on the company's proverbial lips is “Are we covered?" Absent a specific social engineering fraud endorsement, likely not.
Increasingly, insurance carriers are taking the position that unless such an endorsement is attached to a commercial crime policy, the fraud is not covered; not under a cyber policy, as coverage is not triggered when no client data has been stolen; not under a fidelity policy, as the loss may be due to gross negligence or not meet the “lack of knowledge” requirements; and not under a D&O policy, as the loss was intentionally carried out by the employee - notwithstanding the fact that the employee was duped in the process. And often, without the specific endorsement, not even under a commercial crime policy.
Insurers are understandably cautious when underwriting such endorsements in limiting their risk. By some estimates over 100,000 engineering attacks are launched each day1, which explains why sub-limits are often low and deductibles high. Some insurers are even pairing up with private investigation firms to stage a fake president's fraud to assess the implementation of fraud prevention policies actually in place at the company. Others add strict conditions precedent into the policy. Failure to meet them once a fraud has occurred, will likely result in a denial of coverage.
To date, no decisions have been reported in Quebec or the rest of Canada on the coverage of social engineering fraud. However, there have been a number of cases in the U.S. involving fraudulent emails. A notable one is Apache v. Great American Insurance Company2, which involved a fake vendor calling an Apache employee, requesting that payment for services be sent to another account. The employee asked that the request be sent on letterhead. Apache accounts payable then received an email with an attachment on the vendor’s letterhead requesting the change in account information. The employee called the number on the letterhead before entering the change request, and transferring $2.4m. The Court agreed with Apache that this loss resulted directly from computer fraud. The Court held that the term "resulting directly from" is synonymous with a "cause in fact". What is more, a "cause in fact" is established when an act or omission is a substantial factor in bringing about the loss or injury. In this case, the fraudulent email was a substantial factor in bringing about the loss, despite the intervening phone call and supervisor clearance. Therefore, the loss directly resulted from computer fraud. The Court further held that if an insurer could be relieved of paying every time a loss was perpetrated in any way other than a direct hacking, the computer fraud provision would be rendered almost useless. If the insurer wanted to only cover hacking it should have written the policy to reflect this.
Another case currently before the U.S. courts3 involves the classic fake president's fraud scenario described above. In Medidata Solutions, Federal Insurance Company denied coverage under its computer fraud coverage, funds transfer fraud coverage, and forgery coverage. In its complaint, Medidata alleged that computer fraud coverage exists because the altered email address was a fraudulent entry of data or change of data elements directed at Medidata, which constituted a computer violation resulting in a fraudulently induced transfer of money. Medidata also claims that funds transfer fraud coverage exists as Medidata's employees believed that the emails were from a real executive and they became unwitting participants in the fraud. They thereby expressed the instructions of the fraudsters and not those of Medidata. The funds were not sent with Medidata's knowledge or consent since the real executive did not approve this. Medidata also claims that forgery coverage applies since the executive's signature was forged in the fraudulent emails. No decision has been rendered yet.
Not all U.S. courts are siding with insureds. In some cases, they view computer coverage fraud as existing only when someone hacks or gains unauthorized access to a computer system to make an unauthorized transfer of money4. Others take the position that these losses are not directly caused by computer fraud, as it is contingent on the insured's decision to divert the funds from their intended use5.
It will be interesting to see how U.S courts will handle these coverage issues as it is very likely that Canadian courts will be faced with the same questions soon, and may well follow suit.
So what are companies to do to protect themselves? A careful reading of their existing policies should be the first step as some polices may already provide coverage. It is also important to confirm with the insurer or the broker if the company is explicitly insured for that particular type of fraud, and seek the necessary coverage if it is not. Implementation of internal policies to inform all employees that social engineering fraud exists and a step-by-step guide of what to do in the event a dubious email or call is received has also become necessary. To reduce the risk of becoming a target, companies should establish proper countermeasures such as enhancing compliance procedures around financial transfers, establishing call-back procedures, conducting their own third party penetration tests, and avoiding responding to any offers made over the phone or email, to name but a few.
In short, both insureds and insurers would be best served to talk to each other and to others, in an effort to avoid these very costly losses.