Generative AI is becoming one of the hottest commodities on Wall Street, and the social-investing fintech eToro is betting on the tech to improve internal processes and customer-facing products.

Since ChatGPT's splashy debut in November 2022, generative AI has captured people's and corporations' imaginations.

Generative AI is a subset of AI designed to ingest copious amounts of information and create new content. The underlying models, called large language models, can form something new from written prompts. Generative AI has created things as diverse as screenplays, recipes, and romantic partners.

Finance, an industry that runs on data, has been exploring ways to leverage generative AI. At the retail-investing fintech eToro, work is underway to see how generative AI can be deployed to employees and customers.

"As you can imagine, having 32 million registered users and 3 million account holders globally means you got a lot of data. And as you know, AI is a hungry beast that needs data," Lule Demmissie, eToro's US CEO, told Insider. "So it's a natural thing for us to evolve our intellectual property that really runs our Copy investor program to also have AI as part of its DNA."

Demmissie was referring to CopyTrader, eToro's product designed to let users see other people's positions in a social-media-like feed with the option to track and replicate an individual's investing strategies. The company declined to disclose how many investors use CopyTrader.

The fintech said it was experimenting with adding generative AI so users could slice and dice the information in the social feed more deeply and creatively, such as asking to combine two users' portfolios or seeing how many users have large positions in bitcoin.

eToro says employees can expect to benefit from generative AI as well: The fintech's looking at ways to deploy generative AI in its marketing and compliance departments.

In marketing, eToro is testing using AI to translate written content into different languages. Demmissie said that because the company operates in 100 markets, translating marketing materials is "incredibly expensive and complicated when you do that on a manual basis."

In compliance, eToro is exploring ways to present certain disclosures in a fun or interesting way. Demmissie said the fintech had fed "very serious" disclosures subject to Financial Industry Regulatory Authority rules into ChatGPT and asked the AI to make the disclosures sound Shakespearean or like a professor while adhering to regulations.

"Disclosures are boring and dry, and they're important for users to actually understand, so if you have this fun and interesting way of actually communicating these things, it could be an effective way to make sure people pay attention," Demmissie added.

The company said that while it hadn't turned these use cases into products, it was evaluating which generative-AI applications to pursue.

As the company thinks about opportunities that come with generative AI, it can't ignore the risks.

Demmissie said that opening up to generative AI may take a mindset shift, adding that because the tech "is essentially starting to think on its own," mitigating risks would be more about setting up guardrails than controlling the technology.

Demmissie said she was also aware of the models' shortfalls and biases and the moral hazards of dealing with those. For instance, during the disclosures experiment, eToro asked ChatGPT to present the disclosures in a rap.

"The rapper one was terrible! It in no way feels like authentic rap," Demmissie said. "So that sounds superficial, but it's really important that the maker's table culturally includes other people so that people are not left out culturally of the ecosystem."

She added: "I think that's going to be the challenge, because AI will shape our mental framework, it'll shape our narrative, what we feel about ourselves. If the makers don't have all of us, our users, around the table influencing both tone and substance, it's very dangerous."