Goldman Sachs is trialling an autonomous AI software engineer named Devin as part of a broader push to integrate generative AI into its global QA operations, marking a significant milestone in Wall Street’s evolving use of AI in software testing and development, with JPMorgan Chase and Morgan Stanley leading the pack, according to industry observers.
Goldman’s move illustrates how the financial sector’s largest institutions are moving beyond simple productivity tools to adopt agentic AI systems capable of handling complex engineering tasks previously reserved for human developers.
Goldman’s 12,000-strong engineering team is now exploring the potential of Devin to write, review, and debug code, potentially slashing the time spent on routine development work.
A Goldman Sachs spokesperson told various U.S. news channels that the bank is “closely evaluating the productivity gains and guardrails needed to ensure safety and compliance,” highlighting the critical need for oversight in deploying AI tools within tightly regulated environments like banking.
Devin has been developed by Cognition, a San Francisco-based AI startup that recently achieved a $4 billion valuation just 18 months after launch.
The company, backed by venture capitalists including Peter Thiel, made waves earlier this year when it demonstrated Devin autonomously building, testing, and deploying a web application from scratch, without human intervention.
The technology represents a major shift toward agentic AI: systems that not only generate code but execute full workflows, from problem-solving to deployment.
“This is a huge shift,” one tech analyst explained to CNBC. “You give it the prompt, it writes the code, tests it, debugs it, and says, ‘Here you go bank, this is your website’.”
Sector-wide efforts
Goldman Sachs is not alone. Over the past year, JPMorgan Chase, Morgan Stanley and other major financial players have ramped up their internal AI initiatives.
Initially limited to AI-powered email drafting or research, these banks are now pushing further into AI-assisted software engineering. Most systems are built on OpenAI’s large language models, which are integrated into secure environments and fine-tuned with proprietary data to ensure compliance with internal standards and external regulations.
Behind this growing momentum lies a race to modernise the software development lifecycle in financial services, an industry known for its legacy systems and stringent security requirements.
Generative AI offers the promise of speed and scale, but raises new questions about quality control, explainability, and governance.
According to Noah Nzuki, ESG Governance Lead for EMEA at Cognizant, many financial institutions have “embarked on the GenAI journey swiftly but remain at a tactical level.”
Nzuki warned that while experimentation is widespread, the infrastructure for responsible deployment, particularly in the testing and validation of AI outputs, is still “embryotic.”
“Enterprise-wide standards and controls for monitoring, measurement, analysis and model evaluation… are still in early stages,” Nzuki stated, noting that effective GenAI deployment requires robust governance frameworks to avoid risks such as bias, data leakage, or model drift.
The complexity of deploying AI tools in banking is compounded by regulatory and ethical considerations, particularly as these systems become more autonomous.
Human oversight, Nzuki emphasised, remains a key requirement: “It facilitates continuous human input and monitoring, ensuring that AI systems produce accurate and fair results.”
For Goldman and its peers, the goal is not just speed but trust. As agentic AI systems like Devin evolve, banks are grappling with how to ensure transparency and accountability, particularly in high-risk domains such as software testing, which underpins the reliability of customer-facing systems.
The outcome of Goldman Sachs’ pilot may set a precedent for broader adoption across financial services. But for now, the sector stands at a crossroads, balancing the promise of next-gen AI with the foundational need for control, security, and ethical guardrails.
NEW EVENT
Why not become a QA Financial subscriber?
It’s entirely FREE
* Receive our weekly newsletter every Wednesday * Get priority invitations to our Forum events *
REGULATION & COMPLIANCE
Looking for more news on regulations and compliance requirements driving developments in software quality engineering at financial firms? Visit our dedicated Regulation & Compliance page here.
READ MORE
- Leapwork engineering head: Why test automation so often fails to deliver
- World Economic Forum warns financial sector must strengthen AI risk controls
- How Banca Progetto is hard-wiring quality into Italy’s digital banking space
- How Wealthsimple builds quality into the product, not around it
- Digital revamp puts spotlight on internal controls at China Construction Bank
WATCH NOW
