Several banks are adopting agentic artificial intelligence (AI) to enhance their trading oversight, detect misconduct, and heighten compliance efficiency, as reported by Bloomberg News. Deutsche Bank has partnered with Google Cloud to develop AI that can monitor trading activities, identify anomalies, and alert compliance officers for further investigation. This technology will also analyze the communications of client-facing staff, detecting unusual actions such as the transfer of confidential information to personal emails.

Nomura Holdings and another unnamed global bank are also exploring collaboration to train AI models for surveillance, sharing insights, and possibly engaging with regulators. Executives believe such systems could lower false positives by around 40% and decrease compliance costs by $5 million annually.

In addition, Banco Santander is integrating agentic AI into its anti-money laundering (AML) controls with the help of FinTech firm ThetaRay. A report indicated that among U.S. firms with over $1 billion in revenues, 23.4% are either piloting or fully utilizing agentic AI.

Bank of America's CEO, Brian Moynihan, noted that AI and automation are driving productivity improvements across operations. The Federal Reserve is also implementing AI for operational efficiencies in payments, financial management, and services provided to the U.S. Treasury, showcasing a commitment to aligning AI capabilities with specific business needs.

Moreover, Perplexity AI launched its product, Computer, designed to efficiently handle complex assignments autonomously. It can break down extensive tasks, assign subtasks, and manage workflows with minimal human intervention, distinguishing itself from conventional AI models. This offers enterprises enhanced control, accountability, and safeguards while allowing for better performance monitoring—benefitting sectors across the board.

ROCKVILLE, Md., Feb. 19, 2026 /PRNewswire/ -- The rapid evolution of artificial intelligence (AI) is transforming the role of finance leaders, compelling them to integrate innovation with reliability. The 2026 AFP FP&A Forum, set for March 23-25 at the JW Marriott in Indianapolis, will showcase how finance teams are utilizing AI to become strategic business advisors.

Curated by a dedicated task force, the forum features sessions centered around three key pillars that highlight AI's impact on the finance sector. One major benefit is the rise of no-code and low-code platforms, which diminish reliance on extensive IT resources, empowering finance teams to take the lead in AI adoption. This allows them to apply their expertise effectively, turning AI into a vital decision-support tool.

Workshops such as "Designing Finance Bots that Don't Derail the Reporting Cycle" demonstrate advanced automation without requiring coding knowledge. The session "How AI Saved 50 Hours in My Monthly Finance Operations" illustrates a real-world application where one CFO utilized AI to streamline business modeling and improve cash-flow forecasting through rapid scenario analysis.

Moreover, a Fortune 50 agribusiness is leveraging AI and automation to establish a frictionless FP&A ecosystem, showcasing AI's capacity to enhance operational efficiency. The forum also emphasizes the significance of fostering a culture of trust and learning, ensuring that finance teams can thrive in an AI-enhanced landscape.

Ultimately, the sessions at the 2026 AFP FP&A Forum highlight that AI is no longer just a concept; finance teams are embedding it into their core functions to improve analysis, optimize budgeting, and deliver actionable insights for better business decisions. Bryan Lapidus, Director of FP&A Practice at AFP, notes that this integration of AI helps finance professionals elevate their strategic value significantly.

In a pioneering study by Deutsche Bank Research Institute, analysts employed their AI tool, dbLumina, to explore the implications of AI on the global economy. The resulting report, titled “What AI Says About AI Eating Itself and the World,” revealed a significant "great rebalancing," highlighting industries likely to face job displacement due to automation.

Utilizing Google’s Gemini 2.5 Pro model, the AI identified sectors vulnerable to disruption, particularly data-intensive industries reliant on repetitive tasks. Notably, the information technology and software sectors, which are responsible for creating AI technologies, were found to be at risk. With over 85% of developers already utilizing AI coding assistants, productivity gains of up to 60% may streamline operations, but this raises concerns regarding traditional software licensing models and job roles.

The report also targeted the financial sector, especially wealth management, predicting that AI-driven “robo-advisors” could provide the primary financial advice for nearly 80% of retail investors by 2027. In customer service, AI is expected to manage up to 75% of interactions by 2026, relegating human agents to address only complex cases.

Despite its forecast of displacing 92 million jobs by 2030, the AI also anticipates the emergence of 170 million new roles, indicating a net positive outcome for the global workforce. Nevertheless, the transition may lead to substantial disruption, with an estimated 12 million occupational shifts required. While many jobs are deemed vulnerable, positions necessitating human empathy, complex negotiations, and skilled trades remain secure, showcasing the importance of uniquely human skills in a changing job landscape.

Project Spectrum, a partnership between the Bank for International Settlements (BIS), the Deutsche Bundesbank, and the European Central Bank, has ingeniously merged text embeddings with machine learning algorithms to classify millions of products for inflation analysis. This innovative approach was validated using the European Central Bank's Daily Price Dataset (DPD), which consists of billions of daily observations for 34 million unique products.

By transforming raw and fragmented product descriptions into structured data, Project Spectrum empowers analysts and policymakers with real-time, detailed insights into price trends. This capability not only enhances the accuracy of economic modeling but also improves inflation forecasting, marking a significant advancement in AI-driven analysis within economic research.

The project's findings are especially beneficial for monetary policy analysts who rely on high-frequency data for inflation nowcasting. Data scientists at central banks can leverage cost-effective alternatives to traditional large language models for extensive classification tasks. Furthermore, Project Spectrum serves as a valuable technical resource for statistical agencies aiming to automate the categorization of scanner and web-scraped data into official indices. Finally, the methodology established here aids economic researchers in examining price-setting behaviors at the individual product level, fostering deeper insights into market dynamics. Through these concrete applications, Project Spectrum exemplifies how AI can significantly enhance economic analysis and decision-making processes.