Branded content
From agentic AI to tokenised liquidity: How this global bank is engineering the next generation of finance
Powered by people and infrastructure, Citi is contributing to a smarter, faster, more inclusive financial ecosystem through blockchain and tech innovation
By reviewing code and automating routine work with tools powered by artificial intelligence (AI), Citi’s developers can focus more on innovation.
PHOTO: GETTY IMAGES
Follow topic:
With a single prompt on a new agentic artificial intelligence (AI) platform, Citi
Launched in September 2025, the enhanced version of Citi Stylus Workspaces, Citi’s proprietary AI platform, features an upgrade that is now powered by agentic AI.
The latest version is integrated with select Citi systems, enabling employees to complete tasks more efficiently.
Also launched in the same month, Citi Stylus Workspaces’ Agent Mode compresses several manual steps. It is already being piloted by 5,000 users globally and will be rolled out to all employees in the coming months.
Says Citi’s head of application engineering and platforms Payal Jain: “We firmly believe that to be competitive in the digital age, we must be an AI-first workforce.”
Citi was one of the first global banks to introduce AI tools internally in 2024, helping employees summarise documents, draft presentations and automate admin tasks.
The bank has used AI for tasks like fraud detection, for more than a decade – all part of its broader transformation strategy.
To accelerate software development, Citi has leveraged tools such as GitHub Copilot and Cognition’s Devin, as well as the bank’s own Citi Squad.
As of the third quarter of 2025, Citi developers have completed over 1 million automated code reviews and handled routine tasks such as version upgrades and documentation, creating an estimated capacity of 100,000 hours per week and freeing engineers to focus on innovation.
Citi’s agentic AI platform is supporting employees in day-to-day work, simplifying tasks like research, analysis and communication.
PHOTO: GETTY IMAGES
Today, more than 180,000 employees across 83 countries and jurisdictions use Citi AI tools daily, backed by a robust governance framework.
The bank’s ethical AI principles, implemented in 2019, guide every use.
Citi also adopts a human-in-the-loop model when leveraging Citi AI tools to ensure accountability.
“We’ve created channels for colleagues to share what’s working and what they need. That feedback loop is critical to evolving our tools,” says Ms Jain. “Our success hinges on empowering individuals through intuitive tools and fostering an AI-ready culture.”
Currently, more than 4,000 colleagues serve as AI Champions and Accelerators, driving adoption and providing real-time feedback.
There are also mandatory training sessions, such as “Asking Smart Questions – Prompting like a Pro”, and continuous development programmes, says Ms Jain.
“By putting AI into the hands of our people and ensuring they have the skills needed to use the tools effectively, we’re making Citi more productive, agile and client-focused,” she adds.
“We will continue to evolve, leveraging advancements in technology and meeting the needs of our people, our clients and the industry.”
Towards a tokenised future
Besides AI, tokenisation is another force reshaping the financial landscape. Citi has leveraged it to significantly enhance its core services, such as cash management, by using blockchain technology.
These are achieved through Citi Token Services
Explains Ms Ryan Rugg, global head of digital assets for treasury and trade solutions: “Citi Token Services facilitates tokenised internal liquidity transfers and payments, enabling clients to move money around our network, across borders, instantaneously on a 24/7 basis.”
She adds that the platform integrates seamlessly with Citi’s existing products, so clients do not need to adopt new systems.
With its tokenised services, Citi clients can make cross-border payments in real time.
PHOTO: GETTY IMAGES
Citi Token Services for Cash, leveraged on a private permissioned blockchain, went live in 2024, and allows clients to move money 24/7 across Singapore, Hong Kong, the US and the UK, eliminating cut-off times and settlement delays.
For instance, during Chinese New Year this year, a client was able to transfer funds between Singapore and the US while local markets were closed.
In another example, an electronics multinational corporation utilised the service to invest excess liquidity in US money markets after Asia’s close.
“These are real-world examples of programmable payments and instant liquidity,” Ms Rugg notes.
Fueling Singapore’s innovation in digital finance
In Singapore, Citi is a key participant in the Monetary Authority of Singapore’s Project Guardian, a collaborative initiative between policymakers and the financial industry aimed at enhancing the liquidity and efficiency of financial markets through asset tokenisation.
Multiple industry trials have been conducted under the project across asset and wealth management, fixed income and foreign exchange (FX).
In 2025, Citi and Swift proved the feasibility of settling payments between fiat and digital currencies, following successful trials of a Payment-versus-Payment settlement workflow.
The trials demonstrated a powerful hybrid model for interoperability between traditional financial systems and Distributed Ledger Technology networks.
Says Citi’s global head for digital assets for markets Choo Haiping: “We’re proceeding with appropriate caution in this space.
“Our strategy is to pursue opportunities that have technology and regulatory clarity, along with client demand and within our risk capabilities.”
Another pilot was in 2024, when Citi and Fidelity International announced the proof-of-concept of an on-chain money market fund with digital FX swap solution, which demonstrated real-time settlement and improved treasury efficiency.
Mr Choo adds: “This innovation could open potential opportunities to address broader goals, such as portfolio diversification and risk management.”
Learn how innovations at Citi are building a new operating model for finance.
