AI is rewriting finance: Insights from Beijing's Financial Street Forum

Editor's note: Xin Ge, a special commentator for CGTN, is a research fellow at the Institute of Public Policy and Governance, Shanghai University of Finance and Economics (SUFE), and a chair associate professor at the School of Public Administration and Policy, SUFE. The article reflects the author's opinions and not necessarily the views of CGTN.

Artificial intelligence is no longer a subplot in finance; it is becoming central to the industry. That reality will be the subject of the Financial Street Forum (FSF) Annual Conference in Beijing from October 27 to 30. This year's theme – "Global Financial Development in an Era of Innovation, Transformation, and Restructuring" shows how data and advanced analytics is changing the way finance works: how value is created, risk is priced, and competition is won. Organizers expect more than 400 guests from over 30 countries, underscoring the event's role as a convening hub for policy and practice.  

At the heart of this shift is the realization that data itself is now a key asset. Financial institutions that once built mass products for average customers are increasingly able to craft services for a "market of one." A well-known example is Bank of America's virtual assistant, Erica, which has handled over two billion interactions – evidence that continuous intelligence, data-driven service is moving from novelty to norm.

Risk management is undergoing a similar reconstitution. Instead of backward-looking rules and periodic checks, machine-learning systems operating on high-dimensional data streams can surface suspicious patterns in real time. The result is not just more accurate detection but fewer false alarms that drain investigative capacity.

Gains in productivity form the third major change. Automation and AI agents are collapsing cost curves across front, middle and back offices, shifting human effort toward oversight, exception handling, product design and relationship management. Large institutions already report that virtual assistants can resolve a huge number of routine inquiries – which goes a long way in explaining why digital interactions continue to surge at scale.  

Technology alone, however, does not guarantee a healthy change. The FSF's push to forge a "consensus on innovation" acknowledges that shared rules are necessary for progress to build upon itself. Clear explanations, meaningful human guidance and solid testing are essential to avoid spreading bias through large systems. Strong data management – including privacy-preserving techniques and shared learning systems – will be vital to balance performance with increasingly strict data-protection laws. China's own FinTech strategy has focused on speeding up digital change while also improving tech regulation, a framework that will likely shape the forum's discussions.

The forum's program is structured to turn that consensus into practice. According to the conference briefing, FSF 2025 will follow a "Main Forum + Parallel Forums + FinTech Conference + Supporting Activities" format, with 27 thematic sessions and six investment-and-financing matchmaking events. The dedicated FinTech Conference is slated to feature 11 events – evidence that the technology-practice bridge is now central to the agenda.

Crucially, sector regulators will be deeply involved. The People's Bank of China plans to host the Chengfang FinTech Forum on deepening fintech's role in digital and smart transformation, while the China Securities Regulatory Commission's (CSRC) Science and Technology Department will convene a Capital Market FinTech Forum focusing on how artificial intelligence can drive high-quality digitalization across the market. These are less about showcasing and more about working out where specialized technology offers the most benefit, what controls are necessary, and when trial programs should be scaled up.

The event's global reach is also expanding. Organizers note that FSF 2025 will feature five overseas sub-venues – the highest number yet – alongside more participants from international organizations and global financial institutions. That outward posture matters because AI is introducing system-level questions that no jurisdiction can solve alone, from model concentration and data tracking to herding risks as institutions converge on similar tools.

One development to watch is the CSRC's International Advisory Committee, which – according to the regulator – will convene for the first time during the forum. If it becomes a lasting channel for working-level cooperation (like shared incident taxonomies, common validation baselines, coordinated stress scenarios), it could reduce regulatory friction while raising the global benchmark for safety and integrity.  

None of this minimizes the hard problems ahead. The industry still needs common methods to audit model drift, govern third-party dependencies and ensure synthetic data augments rather than distorts reality. Changes to the workforce will require new training so that human-machine cooperation can supplement, rather than supplant, our judgment. But the path is clear. Finance is about to become a fluid, data-driven system for matching risks and resources. Countries that match their technological goals with credible safety measures will be the first to reap the benefits.

FSF 2025 will not settle every debate between October 27 and 30. It can, however, accelerate what matters: scaling the use cases that demonstrably improve inclusion and resilience; hardening governance where risks concentrate; and widening global channels so standards align even as competition intensifies. If that path is taken, the industry will not merely be reshaped – it will be renewed, with technology pushing the limits of what high-quality, safe and inclusive finance can deliver.  

(Cover via VCG)