How ServiceNow is infusing AI everywhere and got 84% of the workforce to use it daily

This unawareness can specifically affect finance processes and the overall finance function. In fact, the old phrase that “to err is human; to really foul things up requires a computer” applies now more than ever. To start with, even the most cutting-edge gen AI tools can make egregious mistakes. Since gen AI can’t do math and can’t “create” out of thin air—instead, it’s constantly solving for a what a human would want—it can “hallucinate,” presenting what seems to be a convincing output but what is actually a nonsense result.

As a financial data company, Bloomberg’s data analysts have collected and maintained financial language documents spanning 40 years. To improve existing natural language processing (NLP) tasks like sentiment analysis, and extend the power of AI in financial services, Bloomberg created a 50-billion parameter LLM—a form of generative AI—purpose-built for finance. About IBM

IBM is a leading provider of global hybrid cloud and AI, and consulting expertise. We help clients in more than 175 countries capitalize on insights from their data, streamline business processes, reduce costs and gain the competitive edge in their industries. IBM’s breakthrough innovations in AI, quantum computing, industry-specific cloud solutions and consulting deliver open and flexible options to our clients.

QuantumBlack Labs is our center of technology development and client innovation, which has been driving cutting-edge advancements and developments in AI through locations across the globe. Later, Lee’s manager asks her to research competitors to analyze their sentiment toward future earnings. Lee uploads the certain sections of MD&A from the latest 10-Q or 10-K for the company and an expanded number of competitors, selected based on Lee’s judgment, into the tool and asks it to analyze information regarding future financial performance. She then prompts the tool to conduct an analysis of sentiment and common themes to determine which company feels more positively about its earnings outlook.

But even the best of the current models have a way to go before they’re ready to fully replace the current crop of smart assistants. Over the past few years, I’ve met several researchers who have used the term “magic” to describe the results of “black box” surrounding large language models. This isn’t a knock against all of the amazing work happening in the space, so much as a realization that there’s still so much we don’t know about the technology. The subject has been a massive question mark looming over Cupertino for the last few years, as competitors like Google and Microsoft have embraced generative AI.

Respondents at AI high performers most often point to models and tools, such as monitoring model performance in production and retraining models as needed over time, as their top challenge. By comparison, other respondents cite strategy issues, such as setting a clearly defined AI vision that is linked with business value or finding sufficient resources. Getting to scale means that businesses will need to stop building one-off solutions that are hard to use for other similar use cases. One global energy and materials company, for example, has established ease of reuse as a key requirement for all gen AI models, and has found in early iterations that 50 to 60 percent of its components can be reused. This means setting standards for developing gen AI assets (for example, prompts and context) that can be easily reused for other cases.

Second, by augmentation—enhancing human productivity to do work more efficiently (such as by gathering and synthesizing multiple pieces of information into a coherent narrative). Third, through acceleration—extracting and indexing knowledge

to shorten financial reporting cycles, and speeding up innovation. Gen AI can greatly enhance CFOs’ ability to manage performance proactively and support business decisions.

Insight Partners backs Canary Technologies’ mission to elevate hotel guest experiences

The EBRI also found that public-sector employers have seen a significant share of their most experienced workers retire or otherwise leave their jobs. For example, BloombergGPT can accurately respond to some finance related questions compared to other generative models. Explore more on how generative AI can contribute to software development and reduce technology costs, helping software maintenance.

Our goal is that this approach, over time, can help shift “Lilli as a product” (that a handful of teams use to build specific solutions) to “Lilli as a platform” (that teams across the enterprise can access to build other products). The broad excitement around gen AI and its relative ease of use has led to a burst of experimentation across organizations. One bank, for example, bought tens of thousands of GitHub Copilot licenses, but since it didn’t have a clear sense of how to work with the technology, progress was slow.

The general idea of staffing squads with resources that are federated from the different expertise areas will not change, but the skill composition of a gen-AI-intensive squad will. While developing Lilli, our team had its mind on scale when it created an open plug-in architecture and setting standards for how APIs should function and be built. They developed standardized tooling and infrastructure where teams could securely experiment and access a GPT LLM, a gateway with preapproved APIs that teams could access, and a self-serve developer portal.

But ensuring the explainability of decisions and actions taken as an outcome of AI algorithms is a complex and multifaceted issue. AI algorithms have dense architecture that relies on numerous parameters and are often an ensemble of interacting models, and whose input signals might not be easily identifiable or even known. Furthermore, there is a general trade-off between model accuracy and flexibility,8 and its explainability. Given the predictive nature of AI/ML algorithms, a key challenge is their ability to minimize false signals during periods of structural shifts. AI/ML models seem to perform well in a relatively stable data environment that produces reliable signals, enabling the models to incorporate evolving data trends without significant loss in prediction accuracy.

A well-defined implementation strategy, supported by training and clear communication, will facilitate a smoother transition, enabling teams to leverage generative AI’s capabilities to the fullest. Integrating generative AI into financial operations demands a meticulous approach tailored to an organization’s needs and objectives. Paramount to this is safeguarding data privacy and security, necessitating stringent measures to counteract evolving cyber threats.

Using biometrics to fight back against rising synthetic identity fraud

This structured format allows for precise, straightforward queries and can easily link related information. Knowledge graphs are used by search engines (like Google’s Knowledge Graph) to enhance search results https://chat.openai.com/ with semantic-search information gathered from a variety of sources. By submitting, you agree that KPMG LLP may process any personal information you provide pursuant to KPMG LLP’s Privacy Statement.

Embracing this technology is crucial to maintaining a cutting-edge finance organization. Learn about Deloitte’s offerings, people, and culture as a global provider of audit, assurance, consulting, financial advisory, risk advisory, tax, and related services. Joshua Henderson is a banking and capital markets research manager in the Deloitte Center for Financial Services.

Many of the most important current opportunities reside outside of the finance function. CFOs should work with their C-suite peers to encourage creative thinking around potential use cases that promote cost efficiency and effectiveness. CFOs can also collaborate with financial planning and analysis and business partners to allocate investments to generative AI and incorporate generative AI-influenced cost targets into the business plan. IT teams will play a pivotal role in prioritizing generative AI investments and addressing data security concerns surrounding the use of AI in finance function applications. Deloitte refers to one or more of Deloitte Touche Tohmatsu Limited, a UK private company limited by guarantee („DTTL“), its network of member firms, and their related entities.

Bias could emerge if the data used to train the system are incomplete or unrepresentative, or the data are underpinned by prevailing societal prejudices. Bias could also arise in the AI algorithm if its design is influenced by human biases. In the financial sector, which is increasingly dependent on AI-supported decisions, embedded bias could lead to, among other things, unethical practices, financial exclusion, and damaged public trust. The deployment of AI applications in the financial sector is raising several concerns about the risks inherent in the technology.

By leveraging AI to automate repetitive tasks and augment human capabilities, businesses can unlock newfound agility and productivity. This not only accelerates operations but also lays the groundwork for scaling AI integration across more complex workflows. The second wave, clearly under way, is analytics empowerment; about half of the CFOs reported that their functions were already using advanced analytics for discrete use cases such as cost analysis, budgeting, and predictive modeling.

Look out for our upcoming perspective on the “Impact of AI in financial reporting and internal controls,” and read about how your company can keep up in this rapidly growing field. One point that quickly becomes apparent when moving forward is that gen AI is not plug and play; companies can’t simply set the models on existing sources of information and let them have at it. Gen AI is a predictive language model—a translator that

sits above existing unstructured data and seeks to generate content that a human would find pleasing.

Legalist, which runs a $1bn hedge fund focused on litigation finance, uses a proprietary AI search tool called “Truffle Sniffer” to find attractive investment targets among a sea of civil suits. Other fund managers are using AI to complement human analysts, identify targets for litigation finance and explain allocation decisions to investors. In March this year, Hedge fund Citadel was navigating to secure an enterprise-wide ChatGPT license.

The prospective implementation envisages bolstering areas like software development and intricate information analysis. “Trends in Employee Tenure, 1983–2022,” a new report by the Employee Benefit Research Institute (EBRI) found that over the past 40 years, the median tenure of all wage and salary workers ages 25 or older has stayed at approximately five years. In 2020 and 2022, the share of workers with the shortest tenure levels increased, while the share with the longest tenure levels also increased.

The analyst inputs a process document and prior credit reviews, including supporting customer information, such as company name, website, and other identifiers.Query. The credit analyst asks the generative AI tool to search for any potential red flags concerning the customer, requesting specific examples of issues such as ongoing legal disputes, business-related concerns, liens, or public disagreements with other vendors.Output. Based on this output and an assessment of the information submitted by the customer, the credit analyst determines that the requested line of credit is acceptable and grants approval. If the tool had identified any red flags, the credit analyst would have needed to validate the information before incorporating it into the final credit decision.

Let’s consider a potential use case—how GenAI might help a junior analyst (Lee) in the financial research and accounting department of a multinational company, during a typical workday. From enhancing trading strategies to fortifying security, Generative AI applications are vast and transformative. However, as with any technology, it’s essential to approach its adoption with caution, considering the ethical and privacy implications. However, as is the case with any evolving technology, generative AI does come with its set of challenges in the finance industry. Up until now, it hasn’t been feasible to incorporate this vast amount of data into a single model due to limited computing resources and less complex/low-parameter models. However, these open-source models with billions of parameters, can now be fine-tuned to large amounts of textual datasets.

Here’s how XAI can help banks confront the “black-box” nature of their machine learning models and develop the tools they need to integrate AI into their business and operations. Generative AI offers seemingly endless potential to magnify both the nature and the scope of fraud against financial institutions and their customers; it’s limited only by a criminal’s imagination. Val Srinivas is the banking and capital markets research leader at the Deloitte Center for Financial Services.

These attacks rely on developing sets of carefully designed prompts (word sequences or sentences) to bypass GenAI’s rules and filters or even insert malicious data or instructions (the latter is sometimes referred to as “prompt injection attack”). Upon its launch on November 30, 2022, Chat Generative Pre-Trained Transformer (ChatGPT) triggered massive global reaction. Remarkably, within a span of two months, the platform gained more than 100 million active users across the globe, a rate much faster than that of other platform innovations (Figure 1).

• Since GenAI can be inaccurate and miss nuance, experienced professionals must oversee and evaluate outcomes.
• ServiceNow, which sells cloud-based software that helps businesses manage their workflows, says the AI investments are already paying off with meaningful gains to ease workloads.
• GenAI is intrinsically geared toward generating new content and using more diverse sets of data sources, it can be used to code synthetic data–generator algorithms, and it better captures the complexity of real-world events.
• Indeed, one of the biggest misconceptions we find is the belief that it’s the job of the CFO to wait and see—or, worse, be the organization’s naysayer.
• A wave of industry spending on chips and data centers to power generative AI is taking priority over corporate spending on AI software from firms like Salesforce and Workday, the Wall Street Journal reports.

Brian’s experience crosses a wide range of industries in the financial services sector, including banking (brokers/dealers), investment companies, business development companies, and alternative funds, including private equity, hedge, and real estate. Brian is also leading Deloitte’s efforts in the Algo/AI assurance area as emerging technologies continue to impact clients and the marketplace. Brian received a BS in Accountancy and BS in Business Administration from Villanova University. It is a large umbrella encompassing many technologies, some of which are already widespread in society and businesses and used daily. When we talk to digital assistants, use autocomplete, incorporate process automation tools, or use predictive analytics, we are using AI.

Developers need to quickly understand the underlying regulatory or business change that will require them to change code, assist in automating and cross-checking coding changes against a code repository, and provide documentation. How a bank manages change can make or break a scale-up, particularly when it comes to ensuring adoption. The most well-thought-out application can stall if it isn’t carefully designed to encourage employees and customers to use it.

Gen AI: A guide for CFOs

These decisions involve developing and marketing products, managing risks, fulfilling regulatory requirements (such as obligations related to anti–money laundering and combating the financing of terrorism), and engaging consumers. Being able to explain Chat GPT financial decisions is at the core of sound financial systems. Annual reports are just one, albeit an important, source that can feed data products. Unstructured data (mostly text) is estimated to account for 80%-90% of all data in existence.

However, the deployment of GenAI in the financial sector has its own risks that need to be fully understood and mitigated by the industry and prudential oversight authorities. As with other technologies, the adoption of generative AI in finance functions will likely follow an S-curve pattern. (See Exhibit 1.) Currently, finance teams are considering how the technology can augment existing processes by creating text and conducting research. Looking ahead, the integration of generative AI will transform core processes, reinvent business partnering, and mitigate risks. Generative AI will eventually collaborate with traditional AI forecasting tools to create reports, explain variances, and provide recommendations, thereby elevating the finance function’s ability to generate forward-looking insights.

Generative AI is part of the new class of AI technologies that are underpinned by what is called a foundation model or large language model. These large language models are pre-trained on vast amounts of data and computation to perform what is called a prediction task. For Generative AI, this translates to tools that create original content modalities (e.g., text, images, audio, code, voice, video) that would have previously taken human skill and expertise to create. Popular applications like OpenAI’s ChatGPT, Google Bard, and Microsoft’s Bing AI are prime examples of this foundational model, and these AI tools are at the center of the new phase of AI.

• GenAI could drive significant efficiency, improve customer experience, and strengthen risk management and compliance.
• For instance, Morgan Stanley employs OpenAI-powered chatbots to support financial advisors by utilizing the company’s internal collection of research and data as a knowledge resource.
• Synthetic data are algorithm-created with a statistical distribution that mimics real data via deep learning model simulation.
• Data leaders also must consider the implications of security risks with the new technology—and be prepared to move quickly in response to regulations.
• While Gemini hasn’t completely conquered Android yet, however, Google is clearly signaling at a day in the not too distant future when it replaces Assistant outright.

Capabilities such as foundation models, cloud infrastructure, and MLOps platforms are at risk of becoming commoditized, given how rapidly open-source alternatives are developing. Making purposeful decisions with an explicit strategy (for example, about where value will really be created) is a hallmark of successful scale efforts. Just as the smartphone catalyzed an entire ecosystem of businesses and business models, gen AI is making relevant the full range of advanced analytics capabilities and applications. Convolutional natural network is a multilayered neural network with an architecture designed to extract increasingly complex features of the data at each layer to determine output; see “An executive’s guide to AI,” QuantumBlack, AI by McKinsey, 2020.

Generative AI might start by producing concise and coherent summaries of text (e.g., meeting minutes), converting existing content to new modes (e.g., text to visual charts), or generating impact analyses from, say, new regulations. Producing novel content represents a definitive shift in the capabilities of AI, moving it from an enabler of our work to a potential co-pilot. AI high performers are much more likely than others to use AI in product and service development. One of the prime use cases suggested by the MIT team is the ability to collate relevant information from these small, task-specific datasets. Tasks include useful robot actions like pounding in a nail and flipping things with a spatula.

AIMultiple informs hundreds of thousands of businesses (as per Similarweb) including 60% of Fortune 500 every month. 5 min read – Software as a service (SaaS) applications have become a boon for enterprises looking to maximize network agility generative ai in finance while minimizing costs. Those realities make it even more important for CFOs to get started in a considered and proactive way. Moreover, company capital (or access to more capital) is finite, and projects compete with one another.

Sooner rather than later, however, banks will need to redesign their risk- and model-governance frameworks and develop new sets of controls. Generative AI (gen AI) burst onto the scene in early 2023 and is showing clearly positive results—and raising new potential risks—for organizations worldwide. Two-thirds of senior digital and analytics leaders attending a recent McKinsey forum on gen AI1McKinsey Banking & Securities Gen AI Forum, September 27, 2023; more than 30 executives attended.

By learning patterns and relationships from real financial data, generative AI models are able to create synthetic datasets that closely resemble the original data while preserving data privacy. AI plays a significant role in the banking sector, particularly in loan decision-making processes. It helps banks and financial institutions assess customers’ creditworthiness, determine appropriate credit limits, and set loan pricing based on risk. However, both decision-makers and loan applicants need clear explanations of AI-based decisions, such as reasons for application denials, to foster trust and improve customer awareness for future applications. We believe that gen AI can have an impact on finance functions in three major ways. First, through automation—performing tedious tasks (such as creating first drafts of presentations).

Cem’s work focuses on how enterprises can leverage new technologies in AI, automation, cybersecurity(including network security, application security), data collection including web data collection and process intelligence. But to that same point of maximizing shareholder value, a CFO must recognize existential threats to a company’s businesses and be clear about the most important levers for generating and sustaining higher cash flows. When an opportunity squarely addresses or significantly relies on gen AI, CFOs should not shunt it aside because they don’t understand the technology or lack imagination to recognize the value it could create. GenAI applications could contribute to liquidity risk if their algorithms inadvertently promote among market participants herd behavior resulting in simultaneous buying or selling decisions; large-scale market dislocations could result.

Generative AI In Finance: Is The Market Ready?

While at present this risk may not be material because current GenAI models are trained and operate on pre-2021 internet scraped data, the situation could quickly change as more people are aware of GenAI capabilities and rapid adoption. Moreover, enterprise-level GenAI applications could be particularly vulnerable, as they use more focused data sets that could be targeted by purpose-built cyberhacking tools. The spread of the use of synthetic data is driven primarily by regulatory necessity and practical business needs. Concerns about data privacy in the context of AI/ML training, particularly in highly regulated sectors like financial and health services, make the use of synthetic data, which cannot be attributed to any person or group, an attractive solution. Synthetic data also offer the opportunity to mitigate imbalances and biases in real data and help build more robust, explainable models that better meet regulatory requirements (Papenbrock and Ebert 2022).

CFOs should select a very small number of use cases that could have the most meaningful impact for the function. In this article, we’ll discuss how CFOs can most effectively approach gen AI company-wide, prioritize specific use cases within the finance function, and rapidly climb the gen AI learning curve. Fine-tuning represents a technique to supplement the training of models like GPT-3, which rely on extensive data sets sourced from diverse origins with more specialized or enterprise-specific data. When the model subsequently generates text, it will produce more focused and accurate output, thereby mitigating the likelihood of spurious or nonmeaningful text.

Other IMF Content

The second factor is that scaling gen AI complicates an operating dynamic that had been nearly resolved for most financial institutions. Just as banks could believe they were finally bridging the infamous divide between business and technology (for example, with agile, cloud, and product operating model changes), analytics and data rose to prominence and created a critical third node of coordination. While analytics at banks have been relatively focused, and often governed centrally, gen AI has revealed that data and analytics will need to enable every step in the value chain to a much greater extent. Business leaders will have to interact more deeply with analytics colleagues and synchronize often-differing priorities. In our experience, this transition is a work in progress for most banks, and operating models are still evolving. Our latest survey results show changes in the roles that organizations are filling to support their AI ambitions.

He supports the firm’s thought leadership initiatives across the industry by generating new ideas, writing insightful publications, and collaborating with practitioners. His current area of research includes the competition of private credit on bank corporate lending. Satish Lalchand is a principal in the analytics practice of Deloitte Transactions and Business Analytics LLP, specializing in anomaly detection and data analytics, business rules development, and modeling. Lalchand has in-depth knowledge of fraud rules creation for prevention, detection, and investigation with a broad range of experience in managing and leading engagements in these areas.

The system is able to combine pertinent information from different datasets into a chain of actions required to execute a task. Because many people have concerns about gen AI, the bar on explaining how these tools work is much higher than for most solutions. So it’s important to invest extra time and money to build trust by ensuring model accuracy and making it easy to check answers. It’s important to bear in mind that successful gen AI skills are about more than coding proficiency. A pure coder who doesn’t intrinsically have these skills may not be as useful a team member. Asset managers are increasingly using artificial intelligence to guide investment decisions, track the habits of portfolio managers and identify moneymaking opportunities.

By now, most companies have a decent understanding of the technical gen AI skills they need, such as model fine-tuning, vector database administration, prompt engineering, and context engineering. In many cases, these are skills that you can train your existing workforce to develop. Those with existing AI and machine learning (ML) capabilities have a strong head start. Data engineers, for example, can learn multimodal processing and vector database management, MLOps (ML operations) engineers can extend their skills to LLMOps (LLM operations), and data scientists can develop prompt engineering, bias detection, and fine-tuning skills. GenAI can be a powerful tool for professionals to more efficiently prepare effective analysis or documentation and enhance their judgments in a variety of areas, including financial planning and research. However, while GenAI can jump-start accounting and financial reporting processes, it still requires a driver at the wheel.

Like all AI, generative AI is powered by machine learning (ML) models—very large models (known as Large Language Models or LLMs) that are pre-trained on vast amounts of data and commonly referred to as foundation models (FMs). To capture the benefits of these exciting new technologies while controlling the risks, companies must invest in their software development and data science capabilities. And they will need to build robust frameworks to manage data quality and model engineering, human–machine interaction, and ethics. Case examples in this article show how these technologies can accelerate and enable access to critical business information, giving human decision makers the information to make thoughtful and timely choices. A successful gen AI scale-up also requires a comprehensive change management plan. Most importantly, the change management process must be transparent and pragmatic.

The users represented a broad spectrum (for example, industries, academia, legal firms, and publishing houses), all of which have started leveraging the technology’s capabilities. By March 2023, several competitors had introduced their own iterations (see Appendix 1) of what are now commonly referred to as generative AI systems (GenAI). Harnessing the power of generative AI requires a large amount of computational resources and data, which can be costly and time-consuming to acquire and manage. Using our AWS Trainium and AWS Inferentia chips, we offer the lowest cost for training models and running inference in the cloud.

Explore how generative AI legal applications can help take actions against fraudulent activities. 3 min read – Generative AI can revolutionize tax administration and drive toward a more personalized and ethical future. Preliminary studies show that over time the biases introduced by GenAI may become even more perpetuated and worse than reality; see Nicoletti and Bass 2023. It typically represents knowledge in terms of entities (like people, places, objects) and relationships between them.

A recent report published by IBM’s Institute for Business Value (IBV) specifies key actions in response to one of seven bets proposed. One action is implementing secure, AI-first intelligent workflows to run your enterprise. You can foun additiona information about ai customer service and artificial intelligence and NLP. It suggests that organizations prioritize which F&A use cases should be augmented with their new foundation models, balancing across precision, risk, F&A stakeholder expectations and return on investment (ROI). This discrepancy between the adoption of generative AI and CFOs’ understanding of it isn’t all that surprising.

Ensure that finance personnel understand how generative AI can complement their work and unlock their potential by automating routine tasks, accelerating business insights, and improving operational efficiency. At the level of the individual analyst, the value proposition includes fewer repetitive tasks and keyboard strokes and more time for business collaboration. An industrial goods company has a prospective customer that requests a line of credit to purchase its products. Because the company does not know the customer, it must conduct a comprehensive credit review before proceeding. The company’s traditional credit review process sought to identify problematic legal or business issues by gathering information from the customer supplemented with additional data collected through third-party sources and internet searches. To expedite the latter task, the credit analyst decides to utilize an internet-enabled generative AI tool.Input.

But scaling gen AI will demand more than learning new terminology—management teams will need to decipher and consider the several potential pathways gen AI could create, and to adapt strategically and position themselves for optionality. The pace of technological advancements means banks won’t fight fraud alone as they increasingly work with third parties that are developing anti-fraud tools. Since a threat to one company is a potential threat to all companies, bank leaders can develop strategies to collaborate within and outside of the banking industry to stay ahead of generative AI fraud. Banks should work with knowledgeable and trustworthy third-party technology providers on strategies, establishing areas of responsibility that address liability concerns for fraud among each party. InScope leverages machine learning and large language models to provide financial reporting and auditing processes for mid-market and enterprises. Read our recently published report, Demystifying algorithms and artificial intelligence, to explore how algorithms are affecting the finance and accounting world.

Future-proofing banks against fraud will also require banks to redesign their strategies, governance, and resources. Looking ahead to the next three years, respondents predict that the adoption of AI will reshape many roles in the workforce. Nearly four in ten respondents reporting AI adoption expect more than 20 percent of their companies’ workforces will be reskilled, whereas 8 percent of respondents say the size of their workforces will decrease by more than 20 percent. That includes the ability to execute tasks that require multiple tools, as well as learning/adapting to unfamiliar tasks.

In fact, 19 other diverse applications are available, each promising to leverage LLMs in novel ways. From prompt engineering to understanding complex financial contexts, FinGPT is establishing itself as a versatile GenAI model in the finance domain. The findings are part of an annual global cross-industry study that surveyed more than 3,000 CEOs from over 30 countries and 26 industries, which included 297 BFM CEOs representing retail, corporate, commercial and investment banks and financial markets. According to a McKinsey report, generative AI could add $2.6 trillion to $4.4 trillion annually in value to the global economy. The banking industry was highlighted as among sectors that could see the biggest impact (as a percentage of their revenues) from generative AI. The technology “could deliver value equal to an additional $200 billion to $340 billion annually if the use cases were fully implemented,” says the report.

These tools and other rules-based innovations are pervasive, but AI is entering a new era. AI is having a moment, and the hype around AI innovation over the past year has reached new levels for good reason. It is transforming from rules-based models to foundational data-driven and language models. With a foundation model focused on predictions and patterns, the new AI can empower humans with advanced technological capabilities that will transform how business is done. These tools include everything from intelligent automation to machine learning, natural language processing, and Generative AI, and they present new opportunities, possible benefits, and many emerging risks for finance and accounting. The data bias problem in GenAI could complicate its adoption and use in financial services.

Data leakages could very well expand beyond private and personal data to proprietary and confidential financial sector data. The initial implementations of these solutions are likely to be aimed internally at financial advisors given that, today, generative AI has limitations with respect to accuracy. In all these cases, the human professional can retain edit rights and final say, and be able to shift focus to other more value-add activities. Generative AI tools can help knowledge workers, such as financial or legal analysts, product innovators, and consultative sales professionals, become more efficient and effective in their roles. © 2024 KPMG LLP, a Delaware limited liability partnership and a member firm of the KPMG global organization of independent member firms affiliated with KPMG International Limited, a private English company limited by guarantee.

Rather than a technical discussion of GenAI, this note seeks to explore the potential risks to the financial sector from this technology based on its current technical characteristics. Generative AI can also rapidly and efficiently produce data products from textual data sources that are only lightly used today. For instance, annual reports and filings (such as 10-Ks filed with the SEC in the United States) are primarily used as a source for financial statements. Buried in text of these documents is data that could power a product catalog or a customer and supply-chain relationship map across all or most public companies globally. Generative AI can create these types of data products at a fraction of the cost that it would take to extract this information manually or with traditional NLP processes. In past blogs, we have described how LLMs can be fine-tuned for optimal performance on specific document types, such as SEC filings.