Major tech companies' agents are collectively entering the heart of the financial sector.

The desktops of finance professionals are becoming a new entry point for major companies to compete for.

On September 3, Tencent officially launched WorkBuddy Financial Edition, bringing together over 80 financial experts and expert teams for institutions such as banks, securities firms, and insurance companies. From due diligence for corporate loans and fund research to customer management for insurance agents, Tencent aims to truly integrate agents into the business processes of financial institutions.

Less than a month ago, Baidu officially named its general-purpose intelligent agent GenFlow "KuKu AI" and launched its own independent office application. Finance was chosen by Baidu as the first key scenario for KuKu AI to transition from general office work to professional office applications.

Earlier, ByteDance's Kouzi had already introduced financial agents and skills from institutions such as Huatai Securities, GF Securities, and Guoxin Securities, bringing market data, financial data, ETF screening, fund comparison and other capabilities into the agent platform.

From ByteDance and Baidu to Tencent, the competition in the general agent market is rapidly penetrating into the financial industry.

It's not hard to understand why large companies are choosing finance. In investment banking, investment research, lending, and insurance, a significant amount of high-cost professional time is still consumed by researching data, reviewing announcements, verifying definitions, updating models, and preparing materials. If agents can take over even a portion of these tasks, they have the potential to directly translate them into efficiency that financial institutions are willing to pay for.

However, the financial market is also a tough one. Data accuracy, traceability of conclusions, and execution authority must all withstand scrutiny. A general-purpose agent that performs well on open networks may not remain reliable once it enters the financial business environment.

Currently, there is no consensus among major companies on whether it is worthwhile to develop a separate agent version for financial applications.

Customer segmentation

Although ByteDance, Baidu, and Tencent have all extended their agents into the financial sector, they are not currently targeting the same market.

Currently, ByteDance's Kouzi and Baidu's KuKu AI are still primarily focused on the consumer (C) side in the financial sector.

Kouzi's approach is to directly integrate professional financial capabilities into its agent. Currently, Kouzi's skill store has launched a series of financial skills from institutions such as GF Securities and Guoxin Securities. For example, GF Securities provides eight skills covering high-frequency investment research scenarios such as financial comparison, stock exchange data, ETF screening, ETF fund flow anomalies, and fund fixed investment.

This means that actions that originally required accessing a brokerage app, financial terminal, or different data pages have been broken down into capabilities that the Agent can directly invoke.

Users simply need to tell Kouzi what they want to research, and it can invoke the corresponding Skill to query market and financial data, compare companies, filter ETFs, or organize fund information. Multiple Skills can also be further combined into continuous tasks for pre-market information organization, intraday monitoring, and post-market review.

Baidu also started with end-users. KuKu AI relies on the content, files, and storage capabilities accumulated by Baidu Wenku, Baidu Scholar, and Baidu Cloud, and has built-in financial data such as listed companies, stock quotes, financial reports, and research reports.

After a user submits a research task, KuKu AI can continue to complete the data retrieval, data processing and content generation, and finally deliver a Word research report, a financial analysis PPT or a financial model Excel. Tasks such as long-term market monitoring can also continue to be executed in the cloud.

Tencent WorkBuddy Financial Edition targets the B2B market: financial institutions such as banks, securities firms, and insurance companies.

In corporate lending due diligence, WorkBuddy can generate a list of materials, identify missing items, cross-verify risks by calling data from industry and commerce, finance, and judiciary, and then form a preliminary due diligence draft based on the bank's internal template. In fund research scenarios, it can complete fund screening and portfolio diagnosis. In insurance scenarios, it can also connect to customer information, product information, and compliance of marketing content.

According to Tencent, since March of this year, WorkBuddy has been introduced to more than 100 financial institutions, including China International Capital Corporation (CICC), SDIC Securities, Ping An Bank, and China Taiping.

However, this application may still be limited to a small scope within financial institutions. According to All-Weather Technology's verification with several securities firms, WorkBuddy and other office agents are not yet widely used internally.

Consensus has not yet been reached

There is no consensus among major companies on whether finance should be developed into a separate agent version.

Judging from the current product forms, apart from Tencent, major companies such as ByteDance and Baidu have not yet launched independent "financial version" agents for financial scenarios.

An insider at KuKu AI told AllWeather Technology that KuKu AI is still positioned as a general office agent, while finance is more often used as a "sample" to demonstrate its ability to perform complex tasks.

The reason is not hard to understand. Finance is highly specialized and involves massive amounts of data. A single task often requires multiple steps, including retrieval, data processing, cross-validation, analysis, and final delivery. If an agent can successfully complete such a complex task in finance, it can more intuitively demonstrate the capabilities of a general-purpose agent.

The aforementioned person stated that KuKu AI will continue to expand its expertise and skills to more industries, rather than focusing solely on finance.

This actually reflects an unresolved issue when office agents move towards vertical industries: should general agents rely on data connectivity, knowledge bases, and skills to continuously expand their capabilities, or should they continue to dig deeper and create a more in-depth product for high-value industries such as finance, law, and healthcare?

The answer is not yet available, but finance is indeed one of the industries that most easily inspires manufacturers to "go down another level".

On the one hand, the unit hourly value of positions such as investment banking, investment research, asset management, and credit is relatively high, but a large amount of time is spent on processes such as searching for announcements, extracting data, updating models, verifying business and legal information, compiling meeting minutes, and preparing materials.

On the other hand, much of the work in finance is already built on digital data. Market data, financial reports, announcements, research reports, business and legal information, and even internal research materials of institutions mostly exist as structured data or electronic documents.

Perhaps it is precisely because of these advantages that WorkBuddy has launched a financial version, creating a separate product for institutions such as banks, securities firms, and insurance companies.

This is also a more aggressive test: if financial institutions are willing to pay for a specialized industry agent, it means that office agents have the opportunity to move beyond selling general productivity tools and enter the market where pricing is based on industry.

However, the financial industry is also challenging because it is not only a high-value industry but also a heavily regulated one.

In June of this year, the State Financial Supervision and Administration Bureau issued the "Guiding Opinions on the Safe Development and Application of Artificial Intelligence in the Banking and Insurance Industries," which listed fund transactions, asset appraisal, credit approval, underwriting and claims settlement, and risk management as high-risk applications of artificial intelligence, and required the establishment of human supervision and intervention mechanisms in key links.

For critical decisions involving customer rights or having substantial financial impact, regulators require the establishment of manual review nodes and the retention of records such as original data and reasoning paths. The document also specifically addresses the need for intelligent agent systems to guard against risks such as data breaches, unauthorized access, tool abuse, and operational malfunctions.

All of this places higher demands on financial agents. On the one hand, finance may be a market where it is relatively easy to calculate the commercial benefits of agent deployment; on the other hand, it is also the industry that first forced vendors to answer questions about data security, model reliability, access control, and the boundaries of responsibility.

Wall Street sets prices first

There is no consensus among major overseas companies on whether finance should be treated as a specialized scenario within a general agent or developed into a separate industry product.

OpenAI leans more towards the former approach, which brings ChatGPT into investment banking scenarios through the Investment Banking plugin, enabling tasks such as company overviews, comparable company analysis, pitchbooks, and due diligence materials; at the same time, it collaborates with PwC to extend the Agent to CFO workflows such as forecasting, reporting, and month-end closing.

Anthropic is also built on Claude's general product system to create financial scenario capabilities.

This year, Anthropic launched 10 financial agent templates covering scenarios such as Pitchbook, KYC, financial modeling, valuation review, and month-end closing. At the same time, it integrated Claude with professional financial data sources such as Excel, PowerPoint, Word, FactSet, S&P Capital IQ, and MSCI.

In contrast, Google chose to go directly to the industry version.

In August of this year, Google launched Gemini Enterprise for Financial Services, specifically designed for capital markets and corporate banking. It features over 50 built-in financial skills and connects to professional data sources such as FactSet, Moody's, MSCI, and PitchBook.

Although the industry has not yet reached a consensus, Wall Street has already priced it in.

In April 2025, financial AI company Rogo completed a $50 million Series B funding round, valuing the company at approximately $350 million. By this year, its valuation had risen to approximately $2 billion, nearly six times its original value in about a year.

Behind the ever-increasing valuation by capital, Rogo truly addresses the disconnect between the basic model and financial work.

Downstream, Rogo connects to professional databases such as Capital IQ and FactSet; upstream, it can generate directly usable Excel models, PPTs, and research reports, while retaining data sources and original citations, allowing analysts to revisit financial statements, conference calls, and investor materials to verify the results.

Rogo currently serves over 300 institutions, covering more than 40,000 financial professionals.

Overall, there is currently no standard answer regarding the ultimate product form of financial agents. It could be a skill within a general agent, a template for a financial agent, or even a standalone industry-specific version like Google and WorkBuddy.

If financial institutions are eventually willing to pay extra for deeper data connectivity, workflow adaptation, and security governance, then financial agents could evolve from today's trial products into an independent software category, and more high-value industries would be re-segmented along this path.

However, if a general-purpose agent can cover most financial needs by relying on stronger models, richer skills, and data connectors, then the space for the so-called "industry version" may be compressed again.

Risk warning and disclaimerInvesting involves risk; please exercise caution. This article does not constitute personal investment advice and does not take into account the specific investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Any investment decisions made based on this information are at your own risk.