Admission Open For 2025-2026

OpenAI Aims to Automate Entry-Level Banking Tasks via Project Codename “Mercury”

Staff Writer: Sweata Maiti

Last Updated: October 22, 2025, 03:09:00 PM IST

Synopsis: OpenAI has initiated a covert project, codenamed Mercury, to train its artificial-intelligence systems to undertake routine tasks traditionally performed by junior investment bankers. The company has enlisted over 100 former bankers from leading investment firms to help build transaction models and prompt frameworks, signaling a strategic push into high-value business applications.

In a bold move to transform the financial-services sector, OpenAI is recruiting ex-investment bankers to help train AI systems capable of replacing much of the repetitive work performed by junior analysts.

Dubbed Project Mercury, the initiative brings on board more than 100 former professionals from financial institutions such as JPMorgan Chase & Co., Goldman Sachs Group and Morgan Stanley. Contractors are reportedly being paid around US $150 per hour to write prompts and build complex models that replicate tasks like initial-public-offering (IPO) modelling, restructurings and financial-forecast decks.

Traditionally, junior bankers spend countless hours assembling Excel-based models and PowerPoint slides, a process often described as the “grunt work” of investment banking. OpenAI’s approach aims to shift this load toward AI systems trained to mimic and enhance those very processes, potentially accelerating deal workflows and reducing labour costs.

According to sources, Project Mercury’s hiring is highly selective. Applicants undergo a multi-stage process beginning with an AI-administered interview, followed by a modelling test focused on financial-statement construction and modelling conventions. Selected contractors produce one model weekly, with feedback loops that feed directly into the AI’s training database.

For OpenAI, this venture is part of a broader strategy to translate its language-model capabilities into enterprise use-cases, even as the firm continues to seek a sustainable business model. Analysts say automating back-office tasks in high-value domains such as banking could pave the way for significant commercial traction.

The initiative also raises questions about labour displacement, especially for junior professionals whose roles may be reshaped or reduced by AI. While efficiency gains are expected, banking-industry insiders caution that fewer entry-level openings may be a by-product of this technological shift.

In sum, Project Mercury signals a notable intersection of generative-AI capability and traditional finance workflows, one that may redefine how routine deal-support tasks are executed in the coming years.