NEW YORK — Citigroup said it is entering a “new phase” in its use of artificial intelligence with the introduction of a platform designed to build and scale AI agents across the organization, while separately, Bank of America is crediting AI and a new tool it has created for its robust earnings.
Citi said its new platform, called “Arc,” will enable the firm to deploy AI agents using what Citi described as the same rigor, discipline and risk framework applied to its broader technology capabilities.
According to Citi, the AI agents are intended to enhance human judgment by taking on tasks such as research, synthesis, preparation and execution, reducing manual effort and accelerating daily operations.
The company cited wealth management as one example, noting that bankers who currently spend hours preparing for client meetings — gathering portfolio data, analyzing market trends and modeling scenarios — could instead rely on teams of AI agents to complete that work proactively and deliver insights when needed. Citi said this shift would allow bankers to focus more on client relationships and personalized service, evolving their role “from coordinator to architect and advisor.”
The Broader Goal
Citi said its broader goal is to reduce operational friction by applying AI to repetitive and manual processes while allowing employees to focus on judgment, sector expertise and client insights.
The launch of Arc builds on what the company described as a strong foundation in AI adoption over the past two years. More than 80% of Citi’s approximately 180,000 employees with access to its AI tools use them regularly, and most have completed prompt training, according to the statement. Citi also said it has modernized infrastructure and strengthened data capabilities to support expanded AI use.
The company emphasized that all AI agents deployed through Arc will be monitored, auditable and governed, allowing the firm to track their activities, performance and value.
Initial Use
Initially, Arc will be used by developers to create agents for specific, well-defined use cases supporting various business lines and functions. Citi said access to the platform will expand over time, with AI agents increasingly integrated into everyday workflows to streamline tasks, surface insights and free up employee time.
Citi said the rollout is part of its broader ambition to become “the most AI-empowered financial institution in the world,” adding that it plans to pursue that goal responsibly.
New Tool at Bank of America
Separately, in Charlotte, N.C., Bank of America reported its strongest earnings in nearly two decades in the first quarter of 2026, with executives pointing to growing use of artificial intelligence—including a new “Meeting Journey” tool—as an increasingly important driver of performance.
The bank said that first-quarter net income totaled $8.6 billion, while earnings per share rose 25% to $1.11, the highest level in nearly 20 years. Speaking on a media call, Chief Financial Officer Alastair Borthwick highlighted the role of AI in improving productivity across the company, particularly within its wealth management business.
A centerpiece of those efforts is the Meeting Journey tool, an internal platform designed to help financial advisors prepare for client meetings more efficiently, according to BofA. The bank has approximately 18,000 advisors serving millions of clients, Borthwick said, and those advisors typically need to review a broad range of information ahead of meetings, including client history, recent account activity and guidance from the firm’s chief investment officer.
Aggregates & Analyzes
The AI-powered tool aggregates and analyzes that information, generating ready-to-use preparation materials. With client consent, it can also function as a notetaker during virtual meetings and produce summaries of key decisions and next steps.
The goal is to reduce the hours advisors spend on manual preparation and allow them to focus more on client relationships, according to the bank.
“Efforts like this translate into results,” Borthwick said, pointing to record first-quarter revenue and improved cost control. He added that while the tool automates much of the data gathering process, human judgment remains central.
“Not necessarily the judgment—that can be human,” Borthwick said.
The company said it invests roughly $13.5 billion annually in technology, including about $4 billion on new initiatives such as artificial intelligence.
Where the Revenue Came From
More broadly, BofA reported its results were supported by strength across several business lines:
- Net interest income rose 9% to $15.9 billion as loan and deposit growth accelerated.
- Trading revenue reached $6.3 billion, its highest level in about 15 years, driven in part by a 30% increase in equities trading.
- Investment banking fees increased 21% to $1.8 billion amid a solid market for mergers and acquisitions.
- Asset management fees climbed 15% to $4.2 billion.
- Productivity gains, including those tied to AI, helped improve the company’s efficiency ratio by 170 basis points to 61%.
Borthwick said the combination of technology investment and operational discipline is helping the bank scale its business while managing costs, with AI playing a growing role in that strategy.
