9 new AI moves in the payments industry

Mastercard expanded its consulting business in late October by launching practices dedicated to AI and economics. It attributed the formation of the AI practice to the "democratization" of generative AI and a complex economic climate, which the card network contends requires new expertise in AI and updated go-to-market strategies. 

"AI has become part of the public consciousness over the past year, and there have been a lot of questions from firms around 'what do I need to do to stay?' " said Les Matthews, executive vice president of services for North American at Mastercard. 

Generative AI is a good fit for Mastercard's diversification strategy, which relies on services, Matthews said. Mastercard, over the past few years, has focused on gaining revenue from consulting, technology and risk management, as regulatory and economic changes pressure card fees. 

The card network uses AI models to provide security for more than 125 billion transactions annually, and is extending that work to other areas tied to AI and machine learning. 

Mastercard's consulting group will identify and integrate uses of generative AI to improve customer experience, treasury management, product testing and general business problems such as improving personalization of communication and reducing bias in credit decisions and marketing. It will also engage with partners such as small businesses to provide education on new forms of AI.

"There's also a lot of confusion over what AI is," Matthews said. "What we're really trying to convey is how we can improve data collection and draw better insights from the data across a broad spectrum of businesses." 

While much of the early use of generative AI across the payments industry has been for internal tasks, such as aiding IT requests or managing internal expense reporting, Matthews envisions external uses such as aiding conversions businesses have with clients and consumers. 

"How do people interact with a website? Or what is their borrowing history?" Matthews said. "The more we can know about what people are doing, the more helpful that is to our partners." 

The payments company has added several new AI tools over the past two months, including some through a partnership with OpenAI, the Microsoft-backed lab that develops ChatGPT. Stripe has connected its payments technology to OpenAI to process the development lab's payments. Stripe's generative AI clients include Moonbeam, which uses generative AI to help content producers write blog posts, essays and other articles. Other clients include Diagram, an AI-powered design company, and Runway, which generates video from text.    

Stripe recently added GPT-4, a new update to ChatGPT, to detect service violations including potential fraud, and to help scale the payment company's support functions.

Stripe referred questions about AI to a CNBC interview with David Singleton, Stripe's chief technology officer. "Stripe has been using AI since the very early days to detect patterns of bad behavior and help protect the millions of businesses that run on Stripe. Today we use machine learning to enhance fraud prevention and detection, but we also aim to use it in various areas across our business to optimize our products," Singleton said, adding that humans and AI will "work together." 

Even with the uncertainty over how generative AI can be used currently, the potential uses are profound, according to Ubaghs. They encompass "everything from marketing campaigns to email templates, and now things like menu design and better integration behind the scenes of a business," he said.

Block introduced 10 generative AI features in late October that sellers can use to automate operations, workflows and other processes. The deployments are part of an initiative launched in March to provide small businesses with new AI tools. The October deployments include an AI menu generator for restaurants and a photo feature that helps businesses add AI-generated backgrounds with images available through a photo studio. Other generative AI tools released in October include personalized email marketing that helps craft subject lines and email content  and a team announcements tool, which helps employers using Square Team Communications to generate announcements to staff via AI-generated copy. 

"A company like Square building these tools directly into their platform, in the context of daily operations, will likely help bridge that connection for many," said Ian Benton, a senior analyst at Javelin Strategy & Research,

Block's clients at launch include Fuddy Duddy, a candle store in Kansas City that is using Block's photo studio app to transition to selling online. The store has used the photo studio, which has created more than 14,000 AI images to replace props and backdrops, saying it has saved hundreds of dollars in marketing expenses. 

"If we can shrink the time it takes a seller to write a website description from 10 minutes to just two or three minutes, that's an important efficiency in the world of small business ownership where the demands are countless and every moment is valuable," Block's public relations office said in an email. 

While ChatGPT's writing ability is impressive, it is also pretty banal, Ubaghs said, adding that's not a bad thing for small businesses.

"It is good at doing mundane and rote writing. It's good at summaries such as technical manuals and FAQs, marketing messages, etc.," Ubaghs said.  "So it's easy to see how a lot of small businesses could make good use of it. The cost savings alone of using GenAI over hiring a menu designer is pretty notable."