Page 32 - GS221202

Education

present one or another offering in the best possible light.
We must expect merchants to be "advised" by increasingly
sophisticated tools that will see through sales pitches that
Legal ease: used to work very well.

Merchant use of payment service pricing analytics raises
interesting questions of title in the underlying data and,
of course, whether the analytics service is itself spinning
the numbers one way or another for its own benefit. Some
ISOs have used the payment service advice model for
AI; 1, lawyers 0 – but years; those same ISOs may now use an analytics bot that
makes the same sale but with seemingly objective data—
that isn't
what will AI do Selling with an algorithm

to payments? Two can play the AI game. At the end of a sale, a human is
often involved in a purchase decision in payments. As we

present us with information that somehow matches our
By Adam Atlas know from social media, AI has the ability to—spookily—
Attorney at Law precise set of interests. ISOs selling payment services
can deploy those AI tools, not just to place advertising at
hatGPT, https://chat.openai.com, is the tip of just the right place and just the right time, but also to go
the iceberg in terms of computer-generated further and pitch a specific merchant as a function of their
content that is strikingly similar to human-gen- subjective interests.
C erated content. I know it will make my profes-
sion obsolete, but what about readers of The Green Sheet? For example, imagine a merchant that already has a
practice of giving some of its profit to a local charity. An
The purpose of this article is to use my vantage point as a ISO selling to that merchant, with the benefit of AI-based
lawyer in payments and crypto to consider how AI might research and pricing, could present that merchant with a
shape the future of payments and the laws surrounding pre-packaged payment portal that provides the merchant
the industry. and their customers with real-time information about the
donations the merchant is facilitating and how they are
Selling to an algorithm changing the community where the merchant operates.

Selling payment services today involves comparing various Selling by way of AI-gathered subjective data points
pricing matrices, sometimes by way of manual crunching might also shift the focus away from pricing and toward
of numbers in spreadsheets. This method of comparison less concrete factors such as community affiliations and
selling and shopping is not likely to last much longer as personal preferences.
merchants will be equipped with AI-driven analytics that
scoops up all of the merchant's data and that of various From a legal perspective, questions will arise as to how far
payment providers to provide points of comparison that a sales organization should or can go in drilling into the
might not have been possible before. psyche of a merchant to make a sale. It's one thing to spray
In this sense, payment providers may have a harder time perfume into the air at a car dealership; it's something
convincing the merchant's algorithm than convincing the else to present every merchant with a unique service that
merchant themselves. For decades, ISOs have leveraged AI has suggested just for them based on their personal
the complexity of merchant processing pricing grids to preferences and data points.

Selling by way of AI-gathered AI is only as good as the data on which it is built. A lot
of legacy data contains our legacy biases. Therefore, from
subjective data points might also a legal perspective, if one were to use AI to tailor sales
pitches, there would have to be real oversight by the
shift the focus away from pricing sales organization to see that the pitches are not simply
and toward less concrete factors perpetuating pre-existing bias in the legacy data, such as
erroneous and harmful differential risk profiles for people
such as community affiliations of different racial profiles.
and personal preferences. Instant contact

Like telling Grandpa he's too old to drive, we lawyers
will soon be told we're too slow and imprecise to write
32