Key Screening Software
Requirements Checklist
IP Address Screening a Must for Providing Online Services/Tech Support

IP Address Screening a Must for Providing Online Services/Tech Support

From last week’s blog, it became clear that using sophisticated screening software that utilizes fuzzy logic is now a necessity to be compliant. Please see:

Two recent OFAC enforcement actions underscore the importance of screening the IP address of a customer to ensure an embargoed country is not at issue. If such a country is implicated, a simple block with a notice, such as “We regret we are unable to provide service for your jurisdiction”, is all that is needed. This was a capability I witnessed while working at HP, Inc. back in 2015 through 2017, but, it turns out, not every company is sophisticated enough to think of this when providing online services/tech support.

The two recent cases are the BitGo and BitPay cases from December 2020 and February 2021 respectively. I will review each case below for conclusions on best practices on IP address screening.

I. The BitGo IP Address Screening Failure

Please see this URL for additional information on this case:

BitGo, Inc. (“BitGo”) is a technology company based in Palo Alto, California that implements security and scalability platforms for digital assets and offers non-custodial secure digital wallet management services. BitGo agreed in December 2020 to remit $98,830 to settle its potential civil liability for 183 apparent violations of multiple sanctions programs.

As a result of deficiencies related to BitGo’s sanctions compliance procedures, BitGo failed to prevent persons apparently located in the Crimea region of Ukraine, Cuba, Iran, Sudan, and Syria from using its non-custodial secure digital wallet management service. All of these countries are under comprehensive U.S. embargo.

Between approximately March 10, 2015 and December 11, 2019, BitGo processed 183 digital currency transactions, totaling $9,127.79, on behalf of individuals who, based on their IP addresses, were located in these sanctioned jurisdictions.

BitGo had reason to know that these users were located in sanctioned jurisdictions based on Internet Protocol (IP) address data associated with devices used to log in to the BitGo platform. BitGo’s reason to know was based on BitGo’s practice of tracking its users’ IP addresses for security purposes related to account logins. BitGo, however, did not use this IP address information for sanctions compliance purposes.

At the time of the transactions, however, BitGo failed to implement controls designed to prevent such users from accessing its services. OFAC determined that BitGo did not voluntarily self-disclose the violations and that the violations constituted a non-egregious case.

II. The BitPay IP Address Screening Failure

Please see this URL for additional information on this case:

BitPay, Inc. (“BitPay”) is a private company based in Atlanta, Georgia that offers a payment processing solution for merchants to accept digital currency as payment for goods and services. BitPay agreed to remit $507,375 to settle its potential civil liability for 2,102 apparent violations of multiple sanctions programs.

BitPay allowed persons who appear to have been located in the Crimea region of Ukraine, Cuba, North Korea, Iran, Sudan, and Syria to transact with merchants in the United States and elsewhere using digital currency on BitPay’s platform. This happened even though BitPay had location information, including IP addresses and other location data, about those persons prior to consummating the transactions.

BitPay’s sanctions compliance program deficiencies enabled persons in these sanctioned jurisdictions to engage in approximately $129,000 worth of digital currency-related transactions with BitPay’s merchant customers. The settlement amount reflects OFAC’s determination that BitPay’s apparent violations were not voluntarily self-disclosed and were non-egregious.

III. Conclusion on IP Address Screening

If a business offers a service online (be it digital currency services, tech support services or other services), it will be necessary to screen the IP addresses of customers prior to providing services. If such customers are from one of the comprehensively embargoed countries (currently, the Crimea region, Cuba, Iran, North Korea and Syria), it will be necessary to impose a transaction block.

The Crimea region of the Ukraine presents a bit of a challenge as one does not want to block all business with the Ukraine. The way we overcame this issue at HP, Inc. is we requested from the U.S. Postal Service all zip codes associated with the Crimea region. Hence, the initial IP address for the Ukraine was flagged, and then, the decision on whether to impose a transaction block was then subject to a secondary flag based on the zip code. If the code was within the Crimea, the transaction block message would be sent and the IP address would be blocked accordingly.

While the BitGo and BitPay penalties may seem relatively low, their cases have been made public, and it is likely that other U.S. person intermediaries (banks in particular) will impose additional due diligence vetting on both of these companies. That, in turn, can cause cash flow delays and outright disruptions. It is therefore well worth undertaking the proper compliance these cases illustrate.

Is Fuzzy Logic Now Clear?

Is Fuzzy Logic Now Clear?

I wrote briefly about “fuzzy logic” in a prior post (please see: What I did not mention is that a couple of America’s biggest companies have recently gotten caught for failures to pick up aliases that their respective software programs should have caught.

The Office of Foreign Assets Control (OFAC) itself has recently upgraded the “fuzzy logic” in its own search tool in January 2021. Please see:

Those two companies that got in trouble recently for inadequate “fuzzy logic” capabilities are Apple and Amazon. This blog will review both cases to illustrate the importance of proper “fuzzy logic” capabilities in a chosen software service provider.

I. Apple’s Failure with Fuzzy Logic

In November 2019, Apple, Inc. agreed to pay $466,912 to settle its OFAC case for apparent violations of the Foreign Narcotics Kingpin Sanctions Regulations. (Please see:

Apple dealt in the property or interests in property of SIS, d.o.o. (“SIS”), a Slovenian software company previously identified on OFAC’s List of Specially Designated Nationals and Blocked Persons (the “SDN List”) as a significant foreign narcotics trafficker (“SDNTK”). The SDN List provided the following identifying information for SIS: SIS D.O.O., 19 Spruha, Trzin 1236, Slovenia; Registration ID 5919070 (Slovenia); Tax ID No. SI91729181 (Slovenia) [SDNTK].

Apple screened the newly designated SDNTKs against its app developer account holder names using its sanctions screening tool. However, Apple failed to identify that SIS, an App Store developer, was added to the SDN List and was therefore blocked. Apple later attributed this failure to its sanctions screening tool’s failure to match the upper case name “SIS DOO” in Apple’s system with the lower case name “SIS d.o.o.” as written on the SDN List. The term “d.o.o.” is a standard corporate suffix in Slovenia identifying a limited liability company.

OFAC determined the following to be mitigating factors regarding Apple’s correction of its “fuzzy logic” deficiency in its screening software, namely, Apple:

• Reconfigured the primary sanctions screening tool to fully capture spelling and capitalization variations and to account for country-specific business suffixes, and implemented an annual review of the tool’s logic and configuration;
• Expanded sanctions screening to include not only app developers, but also their designated payment beneficiaries and associated banks; and
• Updated the instructions for employees to review potential SDN List matches flagged by the primary sanctions screening tool.

II. Amazon Fuzzy Logic Failure, Inc.agreed to pay $134,523 in July 2020 to settle its potential civil liability for apparent violations of multiple OFAC sanctions programs (please see:
As a result of deficiencies related to Amazon’s sanctions screening processes, Amazon provided goods and services to persons sanctioned by OFAC; to persons located in the sanctioned region or countries of Crimea, Iran, and Syria; and to individuals located in or employed by the foreign missions of countries sanctioned by OFAC.

The settlement amount reflects OFAC’s determination that Amazon’s apparent violations were non-egregious and voluntarily self-disclosed, and further reflects the significant remedial measures implemented by Amazon upon discovery of the apparent violations.

Overall, OFAC found the apparent violations consisted primarily of transactions involving low-value retail goods and services for which the total transaction value of the apparent violations was approximately $269,000. OFAC further determined the apparent violations occurred primarily because Amazon’s automated sanctions screening processes failed to fully analyze all transaction and customer data relevant to compliance with OFAC’s sanctions regulations.

What served as chief mitigating factors was that Amazon undertook significant remedial measures to address its sanctions screening deficiencies. Such measures included:

• Employing internal and third-party sources to conduct a thorough review of Amazon’s sanctions compliance program and its automated screening systems in order to address the screening failures that gave rise to the apparent violations. In particular, Amazon is incorporating additional automated preventative screening controls designed to scale and operate effectively for its overall retail business;
• Developing internally custom screening lists to minimize the risk of processing transactions that raise sanctions compliance concerns; and
• Enhancing its sanctioned jurisdiction Internet Protocol (IP) blocking controls and implementing automated processes to update continually its mapping of IP ranges associated with sanctioned jurisdictions.

III. “Fuzzy Logic” Becomes Clear Logic

After reviewing both the Apple and Amazon cases, it becomes readily clear why proper “fuzzy logic” capabilities in picking up alternate spellings and aliases is so important. While such “fuzzy logic” can result in an overwhelming number of “false positive” matches, the logic can be tweaked to match the risk profile of a given company, depending on types of business partners and geographies covered.

Because a given company’s risk profile can evolve over time, it is most helpful to undertake the commitment Apple made in  implementing an annual review of the screening tool’s logic and configuration (likely best done as part of an annual risk assessment).