The meteoric rise of artificial intelligence (AI) over the past year has captured global attention. Tools like ChatGPT and a surge of AI-powered applications have drawn hundreds of millions of users, prompting significant investment from startups and established players alike. This AI boom has extended beyond Big Tech, sparking curiosity within the decentralized finance (DeFi) sector.
Prominent crypto and blockchain-focused investors, including Framework Ventures and Peter Thiel’s Founders Fund, are now backing innovative “crypto + AI” projects such as Sentient and Space & Time. While these ventures primarily aim to disrupt existing AI tech giants, traditional financial institutions have only recently started exploring how AI and blockchain could transform their operations.
Last month, the major oracle protocol Chainlink unveiled an initiative combining AI, oracles, and blockchain technology to address a longstanding challenge: the lack of real-time, standardized data on corporate actions.
A Collaborative Effort by Industry Leaders
The initiative involves a coalition of global financial market infrastructure (FMI) providers, including Swift (the largest interbank messaging platform), Euroclear (a global clearing and settlement firm), investment firms like Franklin Templeton and Wellington Management, and major banks such as UBS, CACEIS, Vontobel, and Sygnum Bank.
Oracles play a critical role in blockchain ecosystems by bridging on-chain and off-chain data. Chainlink, the leading oracle network, has facilitated over $16 trillion in transactions through its data feeds and Cross-Chain Interoperability Protocol (CCIP), which enables seamless communication across blockchain networks.
The Corporate Actions Data Challenge
Financial institutions face significant hurdles due to fragmented corporate actions data. Processes involving mergers, dividends, stock splits, and other events often require coordination among custodians, brokers, fund managers, exchanges, and investors. This journey introduces inconsistent formats, duplicative sources, and even outright errors, as data is exchanged through PDFs, spreadsheets, and press releases.
These inefficiencies result in substantial costs: thousands of regional investors, brokers, and custodians collectively incur $3–5 million annually in manual data validation, with 75% of firms having to re-verify custodian and exchange data.
A Unified Solution with Blockchain and AI
Chainlink’s solution leverages decentralized oracles to create a “unified golden record” of corporate actions data. This record, standardized and ported onto blockchains, can be accessed in real time by all relevant parties, reducing the need for manual reviews.
AI, specifically large language models (LLMs) like OpenAI’s ChatGPT-4, Google’s Gemini 1.5 Pro, and Anthropic’s Claude 3.5, is used to cleanse, validate, and structure financial data before it is integrated into blockchain systems via Chainlink’s CCIP.
Sergey Nazarov, Co-Founder of Chainlink, explained, “Combining AI and oracles turns corporate actions data into highly reliable, structured data. This approach allows asset managers, banks, and FMIs to synchronize data faster, at a fraction of the cost, while minimizing costly errors.”
Transforming Financial Services and Beyond
While the initiative initially focuses on corporate actions data, its implications are far-reaching. Laurence Moroney, an AI researcher and author, noted that Chainlink’s model could be applied to other types of unstructured data, such as legal documents, insurance contracts, real estate agreements, and even social media posts. This could revolutionize industries that rely heavily on unstructured data, enabling better machine readability and streamlined operations.
However, challenges remain. LLMs are prone to “hallucination,” where they generate inaccurate or fabricated data. A study of 11 public LLMs revealed hallucination rates ranging from 3% to 27%. To mitigate risks, pre-trained models tailored to specific datasets are crucial, especially in regulated industries like finance.
Additionally, consensus-based verification by oracles and mechanisms to flag uncertain outputs are essential, ensuring that high-stakes decisions undergo manual review when necessary.
Redesigning Financial Market Infrastructure
For AI and blockchain to achieve widespread adoption, they must address practical, mission-critical challenges within financial institutions. As Stéphanie Lheureux, Director of the Digital Assets Competence Center at Euroclear, observed, the synergy of AI and oracles can “address major pain points and redesign workflows for greater efficiency, transparency, and value.”
This initiative showcases how meaningful innovation often occurs quietly, solving complex, unglamorous problems that underpin global financial markets. Digital financial market infrastructure (dFMI) may not excite the average consumer, but its modernization is essential for creating a more efficient, cost-effective, and accurate financial ecosystem. As we transition into the Web3 era, such advancements will ensure smoother transactions and greater accessibility for businesses and individuals alike.
