Generative AI in Financial Reporting and Accounting Deloitte US

Initiate adoption with use cases whose barriers to entry are low, such as investor relations and contract drafting. Finance personnel will likely find that applying the new technology in real use cases is the best way to climb the learning curve. This iterative approach is essential for cutting through the hype surrounding generative AI and developing a nuanced understanding of the technology’s practical applications and concrete value in the finance function. Generative AI is expected to significantly raise the threat of fraud, which could cost banks and their customers as much as US$40 billion by 2027. Banks should step up their investments to create more agile fraud teams to help stop this growing threat. Regulators are focused on the promise and threats of generative AI alongside the banking industry.

Pilots were conducted last year and generative AI has since been infused into every workplace function ranging from software engineering, HR, customer service, marketing, sales, and financing. Bedi says everyone at ServiceNow is using generative AI today and 84% of the company’s workforce is using the technology daily. While many of the risk issues relating to gen AI are evolutions of discussions that were already brewing—for instance, data privacy, security, bias risk, job displacement, and intellectual property protection—gen AI has greatly expanded that risk landscape.

A social media company’s financial reporting team sends the investor relations team a preliminary draft of the quarterly income statement and balance sheet. Anticipating a strong reaction from the financial markets, the investor relations manager asks an analyst to draft a script for the quarterly earnings call and to formulate potential questions from investors.Input. The analyst imports data from the current and previous quarters into a spreadsheet formatted to be easily understood. To give the tool context and help it understand the types of questions to expect, the analyst also incorporates script drafts and transcripts from previous earnings calls.

These organizations that achieve significant value from AI are already using gen AI in more business functions than other organizations do, especially in product and service development and risk and supply chain management. These organizations also are using AI more often than other organizations in risk modeling and for uses within HR such as performance management and organization design and workforce deployment optimization. Generative AI is a class of AI models that can generate new data by learning patterns from existing data, and generate human-like text based on the input provided.

An AI startup made a hyperrealistic deepfake of me that’s so good it’s scary

Our experts at IBM Consulting are taking a comprehensive look at generative AI for F&A and considering the need to balance risks. AI technology enables finance professionals to focus on higher-value activities, such as strategic planning and analysis, instead of manual and transactional activities. Generative AI empowers faster and better data-driven decisions based on historical data, market trends and the use of AI foundation models that identify patterns and anomalies often missed by traditional analysis methods. Streamlining processes and reducing manual overhead form the crux of this approach.

GitHub reported that developers were working on gen AI projects “in big numbers,” and that 65,000 public gen AI projects were created on its platform in 2023—a jump of almost 250 percent over the previous year. If your company is just starting its gen AI journey, you could consider hiring two or three senior engineers who have built a gen AI shaper product for their companies. The challenge with copilots is figuring out how to generate revenue from increased productivity. In the case of customer service centers, for example, companies can stop recruiting new agents and use attrition to potentially achieve real financial gains. Defining the plans for how to generate revenue from the increased productivity up front, therefore, is crucial to capturing the value.

We have a solid grasp on best practices for training humans how to do different jobs. There are a lot of promising methods, including reinforcement and imitation learning, but future solutions will likely involve combinations of these methods, augmented by generative AI models. 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. Businesses will eventually invest in cloud AI services, but that phase of the AI investment cycle is still a few quarters out, the report says, quoting Wall Street analyst Dan Ives. Following the March earnings quarter, of the 10 largest cloud software providers by annual revenue, WSJ reports eight saw their stocks drop by an average of 9% the day following their earnings reports. 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.

Register to Livestream FHFA’s Generative Artificial Intelligence (AI) in Housing Finance TechSprint

For CFOs to maximize value creation, they must rank the company’s 20 to 30 most value-accretive projects regardless of whether they are AI-related. The Pareto principle always applies; usually a very small number of opportunities will deliver most of the company’s cash flows over the next decade. The CFO cannot let the highest-value initiatives wither on the vine merely because a competing project has “gen AI” attached to it. Sooner or later, shareholders have to pay for everything, and none of them should be on the hook for a gen AI premium. Here, the term flexibility refers to algorithm’s capacity to approximate different functions and is directly related to the number of the model’s parameters.

While the use of gen AI tools is spreading rapidly, the survey data doesn’t show that these newer tools are propelling organizations’ overall AI adoption. The share of organizations that have adopted AI overall remains steady, at least for the moment, with 55 percent of respondents reporting that their organizations have adopted AI. Less than a third of respondents continue to say that their organizations have adopted AI in more than one business function, suggesting that AI use remains limited in scope. Product and service development and service operations continue to be the two business functions in which respondents most often report AI adoption, as was true in the previous four surveys.

It took one financial-services company three months to train its best data scientists to a high level of competence. Leadership put in place cross-functional product teams with shared objectives and incentives to build the gen AI tool. As part of an effort to upskill the entire enterprise to better work with data and gen AI tools, they also set up a data and AI academy, which the dispatchers and service operators enrolled in as part of their training. To provide the technology and data underpinnings for gen AI, the chief data and AI officer also selected a large language model (LLM) and cloud provider that could meet the needs of the domain as well as serve other parts of the enterprise. The chief data and AI officer also oversaw the implementation of a data architecture so that the clean and reliable data (including service histories and inventory databases) needed to build the gen AI tool could be delivered quickly and responsibly. Innovations in machine learning and the cloud, coupled with the viral popularity of publicly released applications, have propelled Generative AI into the zeitgeist.

GenAI could offer a quick and low-cost way for financial institutions to profile their clients, including for risk management, and to screen transactions with a view to identifying the ones that are suspicious. However, the potential of overreliance on GenAI-generated profiles, without appropriate safeguards, could lead to inaccurate or discriminatory client assessments. Appropriate human judgment will need to complement GenAI-based transaction monitoring models. Furthermore, GenAI-based chatbots constitute a particularly sensitive issue when the system Chat GPT is used to address inquiries and complaints by clients, who may not realize they are dealing with an automated system. Nevertheless, the use of chatbots does not excuse a financial institution from its legal and regulatory obligations (see, for example, US Consumer Financial Protection Bureau 2023). The survey results show that AI high performers—that is, organizations where respondents say at least 20 percent of EBIT in 2022 was attributable to AI use—are going all in on artificial intelligence, both with gen AI and more traditional AI capabilities.

Those institutions that successfully harness the prowess of generative AI, while simultaneously respecting its limitations and potential pitfalls, will undoubtedly shape the future trajectory of the global financial arena. Morgan Stanley, for instance, has integrated OpenAI-powered chatbots as a tool for their financial advisors. By https://chat.openai.com/ tapping into the firm’s extensive internal research and data, these chatbots serve as enriched knowledge resources, augmenting the efficiency and accuracy of financial advisory. It currently has 3 versions; the FinGPT v3 series are models improved using the LoRA method, and they’re trained on news and tweets to analyze sentiments.

Employees will not fully leverage a tool if they’re not comfortable with the technology and don’t understand its limitations. Similarly, transformative technology can create turf wars among even the best-intentioned executives. At one institution, a cutting-edge AI tool did not achieve its full potential with the sales force because executives couldn’t decide whether it was a “product” or a “capability” and, therefore, did not put their shoulders behind the rollout. Finally, scaling up gen AI has unique talent-related challenges, whose magnitude will depend greatly on a bank’s talent base. Leading corporate and investment banks, for example, have built up expert teams of quants, modelers, translators, and others who often have AI expertise and could add gen AI skills, such as prompt engineering and database curation, to their capability set.

Given current technological capabilities, the analyst needs to input specific context elements and key insights so that the tool can construct more informed commentary.Query. The analyst asks the generative AI tool to develop a call script (including speaking roles) as well as a preliminary set of likely investor questions and potential responses. He specifically asks the tool to incorporate insights into variances from the previous quarter.Output. The analyst formats the content into a Word document and readies it for an initial review by his manager. To help the CFO prepare, he also highlights the questions most likely to be posed by investors.

It currently excels in text generation and is swiftly honing its skills in numeric analysis. Finance leaders must closely monitor AI’s evolution, gain hands-on experience, and develop their organization’s capabilities. Given the comparatively low entry barriers, there is no need to wait for further advancements before initiating adoption. CFOs should embrace this technology immediately, remove any obstacles to adoption in their departments, and encourage their teams to take advantage of generative AI across the finance function. The findings suggest that hiring for AI-related roles remains a challenge but has become somewhat easier over the past year, which could reflect the spate of layoffs at technology companies from late 2022 through the first half of 2023. While AI high performers are not immune to the challenges of capturing value from AI, the results suggest that the difficulties they face reflect their relative AI maturity, while others struggle with the more foundational, strategic elements of AI adoption.

Banks should be actively participating in the development of new industry standards. By bringing in compliance early during technology development, they can have a record of their processes and systems prepared in case it’s needed for regulators. Banks should focus on their efforts to fight generative AI-enabled fraud to maintain a competitive edge. They should consider coupling modern technology with human intuition to determine how technologies may be used to preempt attacks by fraudsters. There won’t be one silver-bullet solution, so anti-fraud teams should continually accelerate their self-learning to keep pace with fraudsters.

Too often, banking leaders call for new operating models to support new technologies. Successful institutions’ models already enable flexibility and scalability to support new capabilities. An operating model that is fit for scale-up is cross-functional and aligns accountabilities and responsibilities between delivery and business teams. Cross-functional teams bring coherence and transparency to implementation, by putting product teams closer to businesses and ensuring that use cases meet specific business outcomes. Processes such as funding, staffing, procurement, and risk management get rewired to facilitate speed, scale, and flexibility.

Enterprise-level GenAI applications could help mitigate some of the risks inherent in public GAIs, but this option may not be cost efficient for smaller financial institutions. GenAI models could be vulnerable to data poisoning and input attacks (see Boukherouaa and others 2021). Data poisoning attacks attempt to influence AI models at the training stage by adding special elements to the training data set; the effort seeks to undermine training accuracy or to hide malicious actions that wait for special inputs. Input attacks are similar, but they attempt to influence the AI models during operation. Tools, like SEO or GenAI-generated content, could potentially be used to manipulate the GenAI data environment for malicious purposes.

The finance domain can pave the way by establishing an organizational framework that is aligned with your company’s risk tolerance, cultural intricacies, and appetite for technology-driven change. Gen AI can act as an assistant or a coach to employees by helping them do their job more efficiently and ultimately enabling them to focus on strategic, high-impact activities. For example, coding assistance and generation, such as Codey, which is a family of code models built on PaLM 2, can dramatically increase programming speed, quality, and comprehension. Using gen AI can help address some of the most acute talent issues in the industry, such as software developers, risk and compliance experts, and front-line branch and call center employees.

Since GenAI can be inaccurate and miss nuance, experienced professionals must oversee and evaluate outcomes. Professionals may also require training to formulate effective GenAI prompts and guide it to perform a task. In the context of conversational finance, generative AI models can be used to produce more natural and contextually relevant responses, as they are trained to understand and generate human-like language patterns. As a result, generative AI can significantly enhance the performance and user experience of financial conversational AI systems by providing more accurate, engaging, and nuanced interactions with users. 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.

These concerns include embedded bias and privacy shortcomings, opaqueness about how outcomes are generated, robustness issues, cybersecurity, and AI’s impact on broader financial stability. Concerns about risks inherent in GenAI applications are broadly similar to those about AI/ML but with important variations that would need to be considered carefully by the industry and prudential oversight authorities, as detailed below. Furthermore, the distribution of risks between public and private GenAI applications may vary and the risks are likely to be better managed in the latter.

They use an attention mechanism to learn the relationships between parts of the input sequence. This makes them more efficient and easier to train than recurrent neural networks (RNNs). Transformers are better at handling long sequences of data and are also well suited for a variety of NLP tasks. Generative AI has the potential to help financial advisors and investors to leverage conversational text to automatically create highly tailored investment strategies and portfolios.

FinGPT v3.1 is built on the chatglm2-6B model, while FinGPT v3.2 is based on the Llama2-7b model. Developed by AI4Finance-Foundation, FinGPT is currently outperforming other models in terms of both cost-effectiveness and accuracy in general. Bloomberg’s investment of over a million dollars is particularly eye-opening when juxtaposed against the rapid advancements in AI. Astonishingly, a model costing just $100 managed to surpass BloombergGPT’s performance in just half a year. While BloombergGPT’s training incorporated proprietary data a vast majority (99.30%) of their dataset was publicly accessible.

Especially in financial services, where every service or product starts with a contract, terms of service, or other agreement. Gen AI is particularly good at discovering and summarizing complex information, such as mortgage-backed securities contracts or customer holdings across various asset classes. Again, the unstructured nature of much of the data and the size of the data sets add complexity to pinpointing quality issues. Leading banks are using a combination of human talent and automation, intervening at multiple points in the data life cycle to ensure quality of all data. Data leaders also must consider the implications of security risks with the new technology—and be prepared to move quickly in response to regulations.

In some ways, this article is premature—so much is changing that we’ll likely have a profoundly different understanding of gen AI and its capabilities in a year’s time. How well companies have learned those lessons may largely determine how successful they’ll be in capturing that value. But those hoping that gen AI offers a shortcut past the tough—and necessary—organizational surgery are likely to meet with disappointing results. Launching pilots is (relatively) easy; getting pilots to scale and create meaningful value is hard because they require a broad set of changes to the way work actually gets done. So far, Voya’s AI analyst has demonstrated a good ratio of right and wrong decisions, making its alerts “a high-value signal”, said Gareth Shepherd, Voya’s co-head of machine intelligence. He likened the process to a pilot and co-pilot reading signals from an aeroplane’s flight management system.

Other IMF Content

A certified fraud examiner (CFE), he has presented on “Using data analytics” at ACFE and IIA events. The online survey was in the field April 11 to 21, 2023, and garnered responses from 1,684 participants representing the full range of regions, industries, company sizes, functional specialties, and tenures. Of those respondents, 913 said their organizations had adopted AI in at least one function and were asked questions about their organizations’ AI use. To adjust for differences in response rates, the data are weighted by the contribution of each respondent’s nation to global GDP.

Banks spend a significant amount of time looking for and summarizing information and documents internally, which means that they spend less time with their clients. It has sparked excitement around productivity increases and cost savings but also warrants caution.

Said they believed that the technology will fundamentally change the way they do business. The pressing questions for banking institutions are how and where to use gen AI most effectively, and how to ensure the applications are fully adopted and scaled within their organizations. The expected business disruption from gen AI is significant, and respondents predict meaningful changes to their workforces. They anticipate workforce cuts in certain areas and large reskilling efforts to address shifting talent needs. You can foun additiona information about ai customer service and artificial intelligence and NLP. Yet while the use of gen AI might spur the adoption of other AI tools, we see few meaningful increases in organizations’ adoption of these technologies. The percent of organizations adopting any AI tools has held steady since 2022, and adoption remains concentrated within a small number of business functions.

AI/ML may also automate and accelerate the procyclicality of financial conditions through, for example, automating AI/ML’s risk assessments and credit underwriting decisions that are inherently procyclical. In the case of a tail risk event, AI/ML could quickly amplify and spread the shock throughout the financial system and complicate the effectiveness of generative ai in finance the policy response. Earlier this year, Goldman Sachs started experimenting with generative AI use cases, like classification and categorization for millions of documents, including legal contracts. While traditional AI tools can help solve for these use cases, the organization sees an opportunity to use LLMs to take these processes to the next level.

Major technology companies have turned to synthetic data to address a spectrum of operational challenges and objectives. Apple, for instance, employs synthetic data to enhance Siri’s voice recognition capabilities, while Tesla deploys synthetic data in the simulation of myriad driving conditions. Firms in the retail sector are adopting synthetic data to emulate consumer behavior patterns, thereby gaining actionable insights. The launch of ChatGPT has generated fears about the potential risks that GenAI poses. Several major financial institutions have reportedly barred employees from using ChatGPT.2 In April 2023 Italy temporarily banned ChatGPT over concerns regarding potential violations of the European Union’s General Data Protection Regulations.

Climate Change and Sustainability

Today, financial services institutions leverage ML in the form of computer vision, optical character recognition, and NLP to streamline the customer onboarding and know-your-customer (KYC) processes. Generative AI can help firms deliver flexible and relevant conversations that improve the overall customer experience, like adapting the conversational style to match that of the customer (for example, casual conversation mode or formal conversation mode). LLMs can improve employee productivity through more intuitive and human-like accurate responses to employee queries, for example an HR-bot that can answer HR related questions. They also can create more capable and compelling conversational AI experiences for external customer service applications, such as call center assist functionality that provides agents with automated assistance, contextual recommendations, and next best actions.

While demonstrated commercial success has largely come from digital natives, some traditional, nontechnology companies are moving aggressively as well. It is not clear, however, to what extent GenAI could impart some of its risks (for example, bias, accuracy) to the generated synthetic data. If so, this ability will undermine the quality of the synthetic data and their usefulness for training AI/ML systems.

The attractiveness of GenAI for generating synthetic data coupled with the complexity of how the data are generated could potentially blind financial institutions to the potential risks the training data are embedding into their operations. Artificial intelligence (AI) technologies are rapidly transforming today’s business models, and the emerging Generative AI and advanced applications are presenting new opportunities and possibilities for AI in finance and accounting. From Generative AI to machine learning and other foundation model solutions, we look at the new era of AI innovations, the tools they may offer accounting and finance, and considerations for incorporating an AI framework for success.

The pioneering approach optimizes intricate financial strategies and decision-making processes, enhancing efficiency, accuracy, and adaptability in the dynamic world of finance. As the “tip of the spear” in generative AI, finance can build the strategy that fully considers all the opportunities, risks, and tradeoffs from adopting generative AI for finance. As the technology advances, banks might find it beneficial to adopt a more federated approach for specific functions, allowing individual domains to identify and prioritize activities according to their needs. Institutions must reflect on why their current operational structure struggles to seamlessly integrate such innovative capabilities and why the task requires exceptional effort. The most successful banks have thrived not by launching isolated initiatives, but by equipping their existing teams with the required resources and embracing the necessary skills, talent, and processes that gen AI demands.

Finance functions of global companies have not escaped the buzz surrounding the transformative potential of generative AI tools, such as ChatGPT and Google Bard. To see beyond the hype, CFOs need a nuanced understanding of how these tools will reshape work in the finance function of the future. In the short term, generative AI will allow for further automation of financial analysis and reporting, enhancement of risk mitigation efforts, and optimization of financial operations. With its ability to process vast amounts of data and quickly produce novel content, generative AI holds a promise for progressive disruptions we cannot yet anticipate. Access more insights for the banking and capital markets, commercial real estate, insurance, and investment management sectors.

For example, GenAI-generated risk assessment reports based on market sentiments, or customer profile reports from online sources, could be wrong, and this inaccuracy has negative implications for risk-taking and management. Financial services offered to customers through GenAI-supported conversational bots could give inappropriate advice or offer the wrong product to undiscerning clients. These and similar outcomes will expose the financial system to significant risks and erode public trust in AI systems and the financial institutions using them. Bloomberg released training results for BloombergGPT™, a new large-scale generative AI model trained on a wide range of financial domain data.

Similarly, a set of tests for AI/gen AI solutions should be established to demonstrate that data privacy, debiasing, and intellectual property protection are respected. Some organizations, in fact, are proposing to release models accompanied with documentation that details their performance characteristics. Documenting your decisions and rationales can be particularly helpful in conversations with regulators.

3 min read – This ground-breaking technology is revolutionizing software development and offering tangible benefits for businesses and enterprises. These statistics underscore the technology’s role not just in streamlining tasks, but as a strategic asset crucial for financial growth. The business world has always been quick to embrace transformative technologies, from the dawn of the internet to the widespread adoption of cloud computing. However, few innovations have the potential to reshape industries as profoundly as generative AI. We advise CFOs to budget a nominal amount at the learning stage, not for purposes of deploying AI at scale but rather to improve the learning experience for themselves and their team members. Instead, CFOs should select a handful of use cases—ideally two to three—that could have the greatest impact on their function, focus more on effectiveness than efficiency alone, and get going.

• 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.
• It will be interesting how finance team members will “bring you scenarios and forecasts that will require you to engage in the way you didn’t expect,” Schrage told the CFOs.
• Nevertheless, the use of chatbots does not excuse a financial institution from its legal and regulatory obligations (see, for example, US Consumer Financial Protection Bureau 2023).
• For the housing finance system, the transformative potential of generative AI extends beyond technological advancement.

The company hired a chief data and AI officer with a mandate to “enable the organization to create value with data and AI.” The chief data and AI officer worked with the business to develop the strategic vision and implement the road map for the use cases. After a scan of domains (that is, customer journeys or functions) and use case opportunities across the enterprise, leadership prioritized the home-servicing/maintenance domain to pilot and then scale as part of a larger sequencing of initiatives. They targeted, in particular, the development of a gen AI tool to help dispatchers and service operators better predict the types of calls and parts needed when servicing homes. However, enterprise generative AI, particularly in the financial planning sector, has unique challenges and finance leaders are not aware of most generative AI applications in their industry which slows down adoption.