Karl Koster, Noble Systems Intellectual Property and Regulatory Attorney

The FCC is addressing a robocall-related agenda item entitled DECLARATORY RULING AND THIRD FURTHER NOTICE OF PROPOSED RULEMAKING (FCC-CIRC1906-01) at its upcoming Commissioner’s meeting on June 6, 2019. This document would allow carriers to block “robocalls” to their customers, by default, after determining such calls are unwanted and/or illegal. The FCC has stated that “voice service providers may offer opt-out call-blocking programs based on any reasonable analytics designed to identify unwanted calls.”

The analytics algorithms that identify illegal calls or unwanted calls use various criteria. Some analytics providers have stated they use thirty or more criteria. The FCC has listed several criteria in the document, such as call volume patterns, use of invalid numbers, sequential dialing patterns, etc. Each analytics provider considers their blocking algorithms to be proprietary (and likely the best). Some employ graduate level statisticians to constantly improve tune the algorithms. No doubt each would be happy to explain why their analytics are “reasonable.”

Expecting the FCC to define guidelines for evaluating “reasonable analytics” is akin to the Supreme Court trying to evaluate a film against standards for obscenity. As Justice Stewart famously stated, “I shall not today attempt further to define the kinds of material I understand to be embraced within that shorthand description ["hard-core pornography"], and perhaps I could never succeed in intelligibly doing so. But I know it when I see it…”

Trying to quantify and standardize which parameters should be used and how they should be applied to create a “reasonable analytics” blocking algorithm is not likely to be any easier. Rather, the measurement of the effectiveness of a blocking algorithm can be simply accomplished by soliciting the input of the caller and the called party whose call was blocked. If the called party finds that important and wanted calls were blocked, then the blocking algorithm is not effective from their perspective, and would be likely viewed as “unreasonable.” Similarly, if the caller views their calls as legal and learns that their calls are blocked, they would similarly consider the analytics algorithm to be “unreasonable.”

More accurately gauging the “reasonableness” of the algorithms requires a method to inform callers and called parties when a call was blocked. For called parties, this can be easily accomplished by the carrier listing all the blocked calls to the called party on a website. Similar to an email spam folder, the called party could check this periodically to ensure the algorithm is effective. Wanted calls that were blocked could be indicated to the carrier as “wanted” by the called party. Callers could be informed by the carrier routing each blocked call to a network intercept, wherein an announcement informs the caller that their call was blocked. The announcement could inform the caller of a telephone number or website to inform the analytics provider of the alleged erroneous blocking.

The analytics provider could collect this information from the caller and the called party, to fine tune their analytics algorithm. In theory, this input should allow carriers to improve their analytics algorithm over time. In fact, a reducing trend of complaints would support the argument that the analytics were “reasonable.” The FCC should mandate that each voice service provider must report this data quarterly, so that the FCC would know the relative effectiveness of the analytics algorithms from one carrier to another. A carrier with a relatively high number of complaints would be an outlier, and their analytics algorithm could be viewed as being “unreasonable.”

If carriers are allowed to block calls in the manner outlined by the Commission, it should mandate both callers and called parties be informed about each call being blocked, and both should have a means to inform the carrier that a mistake was made. These inputs will serve as useful data to the Commission in evaluating whether the analytics are “reasonable” and will avoid the Commission from attempting to address the proprietary operation of each analytics algorithm. No doubt if Justice Stewart were to use this basis to opine as to whether an analytics algorithm was reasonable, he would then “know it when I see it.”

The opinions presented here are those of Karl Koster, and not necessarily those of Noble Systems. The contents should not be construed as legal advice nor as comments reflecting any regulatory position of Noble Systems. Comments can be sent to kkoster@noblesystems.com.