We have all been there. You have a customer service issue. It should take 30 seconds to explain, so you call the company that sold you a good or provided you with a service that wasn't quite right. You're confident that a quick phone call will get the problem sorted. Then it happens - your call connects and you're greeted by an almost human sounding voice asking what help you need and prompting you for information to get the process started. It is at this point where my Irish temper starts to percolate. Why? Because experience informs that I am at the beginning of an automated customer service labyrinth that will leave me exasperated and - too often - without resolution of the issue that prompted my call.
Stepping up to the plate to address these issues, and perhaps lower my blood pressure, is the New York City Department of Consumer and Worker Protection (DCWP). The DCWP has issued a proposed rule to address the deceptive use of "automated customer services systems." It would be situated in the Unfair Trade Practices provisions of the New York City Rules and Regulations. So, what will this rule require, and what does it mean for the financial services industry? Let's dig in.
Under the proposed rule, an "automated customer service system" is any form of technology used to provide consumers with "customer service" where all (or substantially all) of the communication with the consumer is made using technology instead of a human being. The types of "customer service" inquiries covered by the proposed rule is broad. It would cover any assistance that a "merchant" provides in response to a consumer's inquiry about a product, service, account, refund, billing matter, a complaint, a consumer's request to cancel, a fraud claim or a dispute.
One potential point ripe for future revision is the absence of a definition of the term "merchant." This term is embedded in the definition of "customer service" but is not defined either in the proposed rule or elsewhere in the unfair trade practices provisions. As a result, the general meaning of the term will apply. The Merriam-Webster Dictionary defines "merchant" to mean the "operator of a retail business." This seems broad enough to cover a financial services provider. Moreover, because the proposed definition of the term "customer service" notes that it will include inquiries regarding both products and services, it seems clear that the rule would govern the financial services industry.
So, if the proposed rule does apply to the financial services industry, then what does it require? The proposed rule approaches the regulation of automated customer service systems via disclosure. For starters, it would make it a deceptive trade practice for a merchant to represent that customer service is available through a particular communication channel (for example: a telephone call or via email) if that is not the case. This seems straightforward: don't promise something that you don't offer. The rule clarifies that this prohibition would not be violated by general consumer-facing statements (on a website, for example) that provides contact information unless such contact information is accompanied by terms such as "customer service" or "contact us" or similar language that connotes the point of contact is one the consumer may use for customer service inquiries.
Coupled with this prohibition, the rule would also require merchants using automated customer service systems in a "covered interaction" to provide an up-front disclosure of that fact to consumers. A "covered interaction" is defined, in somewhat circular fashion, to mean any interaction between a consumer an automated customer service system. In short: merchants would be required to tell consumers at the beginning of a customer service contact that the consumer is interacting with an automated customer service system rather than a human representative. And this requirement comes with a language access requirement, as the disclosure would need to be in the same language used in the interaction.
The rationale for this disclosure is found in the Statement of Basis and Purpose for the proposed rule. It posits that consumers who know from the beginning of an interaction that they're talking with an automated customer service system will modify their language and set their expectations appropriately. However, the Statement of Basis also cites to studies that show that nearly three quarters of consumers who interacted with a chat bot described the experience as "a complete waste of time," and that the number of consumers who found AI-powered customer service to have declined by 10% since 2021 - precisely during a period of rapid involvement of such systems. With those data points in mind, one wonders if this disclosure will prompt consumers to simply hang up or otherwise abandon the interaction altogether.
The proposed rule would also make it a deceptive and unconscionable trade practice for a merchant to fail to provide customer service it in a "reasonable and timely manner." This would arise if the merchant engages in conduct that "repeatedly and unreasonably" directs consumers to menus that fail to adequately address the consumer's inquires or that respond provide "unresponsive, irrelevant, or erroneous information." Sound familiar? The rule would make it a rebuttable presumption that this standard is violated if the merchant uses an automated customer service system when interacting with consumers unless it provides consumers the option to interact with a human representative as dictated by the rule.
One means of rebutting this presumption would be to provide a consumer with an eject button: a way to break away from the automated interaction and talk to an actual person. The proposed rule states that if the consumer can connect to a human customer service representative in a "single-step transfer" (by saying or typing "agent" or pressing "0" for the operator) at any point during the interaction, and be told of the estimated wait time, the process will pass muster. Such transfers must be a clearly identified and easy to use (for example: by hitting a button or using a voice command or menu selection). The rule would prohibit requiring consumers to complete any additional automated steps, answer any additional questions, provide additional justification for the transfer, or otherwise require the navigation of additional menus or processes before making the transfer unless such menus or processes are needed to route the consumer to the appropriate human customer service representative (so the "I'll just need a few more pieces of information" inquiries may not be a thing of the past).
What if the interaction of the consumer and the automated customer service system happens outside of business hours when a human representative is not available, or when hold times might be unreasonably long? The eject button for the consumer comes with a parachute for the merchant. The proposed rule clarifies that in such instances the merchant may comply by providing the consumer with instructions for how to access an alternative means to obtain assistance from a human representative. The alternative means must include at least one of the following:
While not insurmountable, each of these options would need to be compared to current systems infrastructure to determine their feasibility, timing and cost. For example, the drafting of the "eject button" portion of the rule (the definition of "single-step transfer mechanism," specifically) states that this option must be something the consumer is able to use at any time. This would appear to prohibit systems that would only allow this option at the end of the segment of the interaction between the consumer and the automated customer service system. It is unclear whether current systems would be able to accommodate this requirement.
The proposed rule would also impose recordkeeping requirements, related to both the automated customer service system used and its implementation, but also regarding its functioning. Merchants would be required to retain logs detailing information of consumer interactions including the duration of the interactions, whether a single-step transfer mechanism was offered and used by the consumer and whether the consumer was transferred to a human representative, placed into a queue or provided an alternative means to interact with a human representative. Merchants would also be required to compile a variety of aggregate information on covered interactions on at least an annual basis.
Finally, the proposed rule comes with sharp teeth. It proposes a tiered penalty structure for violations of the law ranging from $ 525 to $ 3,500 per violation. Higher amounts would be assessed for recidivist activity. In a city with a population larger than most states, and the number of times these types of interactions may occur on a daily basis, the numbers theoretically become eye-popping.
Whether this proposed rule makes it to the finish line in its current form or at all is, at this point, unknown. The public hearing for, and deadline for comments regarding, this proposed rule was initially scheduled for November 19th, but was recently pushed back by a month to mid-December. As such, financial service providers who currently use automated customer service systems, or who are thinking of deploying one, should review the rule and pay particular attention to what it would require against current operations. And when doing so users of such systems should also be mindful of the fact that a handful of states with laws governing human interactions with technology of this sort (California, New Jersey and Utah come to mind). As artificial intelligence system improve and further blur the lines between what is flesh and blood and what are chips and wiring these types of consumer protection rules may only increase in number.
Thomas P. Quinn, Jr., is a partner in the Massachusetts office of Hudson Cook, LLP. Tom can be reached at 774.365.4758 or by email at tquinn@hudco.com.
Copyright © 2026 CounselorLibrary.com, LLC. All rights reserved.