S&P completes ‘operational integration’ of IHS Markit, ups focus on genAI, new regtech tool

After piloting a new generative AI-powered search interface, ChatIQ, in December to a handful of customers, S&P Global is planning a large swath of similarly AI-led new products and enhancements to existing products, president and CEO Doug Peterson told investors during the company’s quarterly earnings call on Thursday. 

Later this year, S&P expects to launch RegGPT, for which the company last year registered a trademark, which is still pending. The product includes downloadable computer software for the artificial generation of human speech and text; natural language processing, generation, understanding and analysis; machine learning-based language and speech processing; simulating conversations; and creating and generating text, all in the field of regulatory compliance, according to the application filed with the US Patent and Trademark Office.

S&P also recently added a genAI search capability to Marketplace, its data “storefront” through which users can browse and purchase datasets sourced from all of S&P’s business segments.

“We’re using generative AI to allow people to query into those datasets to get broader understanding of what’s actually available through our company. We’re pretty excited about that, and early responses have been positive,” said Adam Kansler, president of S&P Global Market Intelligence. In addition to RegGPT, the company is also planning to release “some other tools around our ratings-related data that we deliver through Market Intelligence to allow our customers to probe into that data and get quick responses that help increase the speed within their workflows.”

We believe that AI will quickly become embedded in everything we do

Doug Peterson, president and CEO, S&P Global

As further commitments to its AI strategy, S&P, in conjunction with its Kensho business—which S&P acquired in 2018 for $550 million—has also developed an AI accelerator to fast-track high-priority AI initiatives and build common capabilities that can be deployed across the company.

“We believe that AI will quickly become embedded in everything we do,” said Peterson. But he acknowledged a challenge that even data-rich companies face, which is that much of the company’s data isn’t yet ready to be ingested by large language models (LLMs). To combat this, S&P again intends to leverage Kensho for data cleaning, data transformation, data reduction, data integration, tokenization, and tagging through its proprietary tools: Scribe, Nerd, Link, Extract, and Classify.

“Having so much AI expertise in-house means that we can leverage infrastructure and compute platforms from multiple hyperscale cloud providers, third-party LLMs, our own proprietary LLM, and a wide array of other vendors without having to lock ourselves in or cede economics to one ecosystem or another,” Peterson said. “Ultimately, the goal is to have generative and traditional AI capabilities embedded everywhere that makes sense.”

S&P utilizes cloud offerings from Amazon Web Services, Microsoft Azure, and Google Cloud, though in February of last year, the data giant signed a multi-year strategic collaboration agreement with AWS. As part of the deal, AWS became S&P’s primary cloud provider, Kansler told WatersTechnology last year.

Peterson also spoke about the merger with IHS Markit, which formally began nearly two years ago. As the second anniversary approaches, Peterson said that S&P can “definitively say it has been a success” and that the company has “put the work of operational integration behind us.”

After identifying cost and revenue synergies, Peterson said the combined company is now starting to deliver new products, offering an example from the Commodity Insights division called Platts Connect, a blending of the S&P Platts Dimensions Pro and IHS Connect into a new platform for prices, research, and forecasting.

“Over the course of 2023, we launched seven new products just within Market Intelligence. We have 14 new products set to come to market across 2024,” Kansler said. “These will increasingly become part of what we’re offering out to customers. Whether it’s just putting our bond pricing in together with our credit analytics capabilities, looking at our economics and country risk desktop across our desktop, incorporating sustainability data into our private markets portfolio management capabilities—these are all exciting areas.”