Around 18 months ago, Meta launched its mostly free open-source Llama models, which have since seen nearly 350 million downloads. A number of major firms have adopted them since then, including certain financial services.

Now, Goldman Sachs and Nomura joined the ranks of financial companies that started using the Llama models, as revealed by Meta’s recent progress report. The report revealed that Goldman Sachs’ GS AI Platform allows the bank’s engineers to use the Llama models for a variety of different use cases.

One example is extracting information from documents, which is done much faster through the use of artificial intelligence than manually.

As for Nomura, the global financial services group with an integrated network that includes over 30 countries and regions, the firm uses Llama on AWS. The idea is to achieve much faster innovation, bias guardrails, transparency, and performance across different areas, including code generation, text summarisation, document processing, and log analysis.

Meta does not regret taking the open-source route with AI models development

Meta, Facebook’s parent company, has invested billions of dollars into the artificial intelligence space, firmly believing that this is the next step in innovative technologies that will spread around the world. However, it is also worth noting that it has taken a different approach to other major names in the AI space, such as OpenAI — at least when it comes to its open-source models.

The firm’s founder and CEO, Mark Zuckerberg, published an open letter this July, arguing that open-source AI is good for Meta to focus on. He noted that this technology can prevent the company from getting locked into a closed ecosystem belonging to one of its competitors.

With its own AI models, Meta is free to innovate, experiment, and evolve on its own. Zuckerberg also added that the bottom line is that open-source AI represents the world’s best shot at harnessing this technology to create the greatest economic opportunity, and provide security for everyone.

While some have argued that it was a risky move for Meta to take the open-source AI development route, the firm’s chief executive argued that open-source development of AI would allow technologists to learn of the power of AI models. With this knowledge, they could create their own AI, as well.

Zuckerberg also said that it is unrealistic to think that a handful of firms could keep their AI tech secret, especially since Silicon Valley has been a target for espionage by nations such as China for years now.

“I think governments will conclude it’s in their interest to support open source because it will make the world more prosperous and safer,” he argued at the time.