Using AI to Improve Customer Experiences in Finance Deloitte US

The Task Force is currently conducting a strategic Review of the Principles to identify new or emerging developments in financial consumer protection policies or approaches over the last 10 years that may warrant updates to the Principles to ensure they are fully up to date. The Review will include considering digital developments and their impacts on the provision of financial services to consumers. The G20 Riyadh Infratech Agenda, endorsed by Leaders in 2020, provides high-level policy guidance for national authorities and the international community to advance the adoption of new and existing technologies in infrastructure.

Certain services may not be available to attest clients under the rules and regulations of public accounting. Use conversational AI solutions such as chatbots and virtual assistants to handle a wide range of consumer-facing activities — from helping consumers find a better credit card or cancel unneeded accounts, to negotiating collections. In conclusion, although the race of building large-scale models isn’t slowing down, a new rising trend of data-centric AI is receiving wider recognition and resonance from the AI community. While powerful models are exciting, if there isn’t enough of the right data to run these models, the advancements are stunted. Furthermore, data-centric AI supports compliance with regulatory considerations to control AI application in finance. Moreover, generative AI models can be used to generate customized financial reports or visualizations tailored to specific user needs, making them even more valuable for businesses and financial professionals.

Principle 7: Protection of Consumer Assets

As we can see, the benefits of AI in financial services are multiple and hard to ignore. According to Forbes, 65% of senior financial management expects positive changes from the use of AI in financial services. Employing robotic process automation for high-frequency repetitive tasks eliminates the room for human error and allows a financial institution to refocus workforce efforts on processes that require human involvement. Ernst & Young has reported a 50%-70% cost reduction for these kinds of tasks, and Forbes calls it a “Gateway Drug To Digital Transformation”. The predictions for stock performance are more accurate, due to the fact that algorithms can test trading systems based on past data and bring the validation process to a whole new level before pushing it live. Unfortunately, it’s common for AI models to undergo training using biased datasets that may underrepresent certain groups of people.

Specific software, such as enterprise resource planning (ERP,) is used by organizations to help them manage their accounting, procurement processes, projects, and more throughout the enterprise. Examples of back-office operations and functions managed by ERP include financials, procurement, accounting, supply chain management, risk management, analytics, and enterprise performance management (EPM). AI assistants, such as chatbots, use AI to generate personalized financial advice and natural language processing to provide instant, self-help customer service. Alpaca uses proprietary deep learning technology and high-speed data storage to support its yield farming platform. (Yield farming is when cryptocurrency investors pool their funds to carry out smart contracts that gain interest.) Alpaca is compatible with dozens of cryptocurrencies and allows users to lend assets to other investors in exchange for lending fees and protocol rewards.

Explainability

The scaling up of the use of algorithms that generate uncorrelated profits or returns may generate correlation in unrelated variables if their use reaches a sufficiently important scale. It can also amplify network effects, such as unexpected changes in the scale and direction of market moves. Machine learning is the subset of AI that focuses on building systems that learn—or improve—performance, based on the data they consume, without necessarily requiring various human interventions, such as programming and coding. Shapeshift is a decentralized digital crypto wallet and marketplace that supports more than 750 cryptocurrencies. The platform provides users access to nine different blockchains and eight different wallet types. ShapeShift has also introduced the FOX Token, a new cryptocurrency that features several variable rewards for users.

At the same time, the deployment of AI in finance gives rise to new challenges, while it could also amplify pre-existing risks in financial markets (OECD, 2021[2]). AI algorithms can analyze transactions in real time, detect anomalies and patterns that may indicate fraudulent activities, and alert banks to take appropriate actions. PayPal uses machine learning algorithms and rule-based systems to monitor real-time transactions, and identify potentially fraudulent activities.

Customer service

Although a convergence of AI and DLTs in blockchain-based finance is promoted by the industry as a way to yield better results in such systems, this is not observed in practice at this stage. Increased automation amplifies efficiencies claimed by DLT-based systems, however, the actual level of AI implementation in DLT-based projects does not appear to be sufficiently large at this stage to justify claims of convergence between the two technologies. Instead, what is currently observed is the use of specific AI applications in blockchain-based systems (e.g. for the curation of data to the blockchain) or the use of DLT systems for the purposes of AI models (e.g. for data storage and sharing). The difficulty in comprehending, following or replicating the decision-making process, referred to as lack of explainability, raises important challenges in lending, while making it harder to detect inappropriate use of data or the use of unsuitable data by the model. Such lack of transparency is particularly pertinent in lending decisions, as lenders are accountable for their decisions and must be able to explain the basis for denials of credit extension.

Research suggests that explainability that is ‘human-meaningful’ can significantly affect the users’ perception of a system’s accuracy, independent of the actual accuracy observed (Nourani et al., 2020[42]). When less human-meaningful explanations are provided, the accuracy of the technique that does not operate on human-understandable rationale is less likely to be accurately judged by the users. In the most advanced AI techniques, even if the underlying mathematical principles of such models can be explained, they still lack ‘explicit declarative knowledge’ (Holzinger, 2018[38]).

The Growing Impact of AI in Financial Services: Six Examples

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. Aggregators like Plaid (which works with financial giants like CITI, Goldman Sachs and American Express) take pride in their fraud-detection capabilities. Its complex algorithms can analyze interactions under different conditions and variables and build multiple unique patterns that are updated in real time. Plaid works as a widget that connects a bank with the client’s app to ensure secure financial transactions.

• Section three offers policy implications from the increased deployment of AI in finance, and policy considerations that support the use of AI in finance while addressing emerging risks.
• Socure created ID+ Platform, an identity verification system that uses machine learning and AI to analyze an applicant’s online, offline and social data, which helps clients meet strict KYC conditions.
• —Federal Reserve Chair Jerome Powell said of Silicon Valley Bank in a press conference following a Fed decision to hike interest rates 0.25%, Yahoo Finance reported.
• Evidence based on a survey conducted in UK banks suggest that around 35% of banks experienced a negative impact on ML model performance during the pandemic (Bholat, Gharbawi and Thew, 2020[50]).
• Distributed ledger technologies (DLT) are increasingly being used in finance, supported by their purported benefits of speed, efficiency and transparency, driven by automation and disintermediation (OECD, 2020[25]).

Deep learning algorithms are capable of processing huge amounts of data including stock prices and trading volumes. This technology is also being used to develop trading algorithms that can buy and sell automatically based on market data in real-time. The technology of natural language processing involves the analysis and generation of human language. Generative AI is an emerging technology that is starting to gain traction in the finance industry. Generative AI can be used to create synthetic data that mimics real-world financial data, which can be used to train machine learning models to recognize patterns, identify trends as well as make predictions.

Currently, financial market participants rely on existing governance and oversight arrangements for the use of AI techniques, as AI-based algorithms are not considered to be fundamentally different from conventional ones (IOSCO, 2020[39]). Model governance best practices have been adopted by financial firms since the emergence of traditional statistical models for credit and other consumer finance decisions. Documentation and audit trails are also held around deployment decisions, design, and production processes. It allows financial institutions to overcome data limitations, improve risk management, enhance fraud detection, personalize services, and refine investment strategies.

Use AI and machine learning to detect transactional and account takeover fraud across the banking value chain. Scaling AI across financial organizations means addressing the challenges of data silos, internal departments, industry regulations, and data protection. As a fine-tuned generative model for finance, it outperformed other models by succeeding in sentiment analysis. This automation not only streamlines the reporting process and reduces manual effort, but it also ensures consistency, accuracy, and timely delivery of reports. – By 2025, 70% of organizations will use data-lineage-enabling technologies including graph analytics, ML, A.I., and blockchain as critical components of their semantic modeling. Intelligent character recognition makes it possible to automate a variety of mundane, time-consuming tasks that used to take thousands of work hours and inflate payrolls.

Credit and Risk Underwriting

In addition to the inherent complexity of AI-based models, market participants may intentionally conceal the mechanics of their AI models to protect their intellectual property, further obscuring the techniques. The gap in technical literacy of most end-user consumers, coupled with the mismatch between the complexity characterising AI models and the demands of human-scale reasoning further aggravates the problem (Burrell, 2016[37]). The Policy Guidance supports the development of core competencies on digital financial literacy to build trust and promote a safe use of digital financial services, protect consumers contribution margin from digital crime and misselling, and support those at risk of over-reliance on digital credit. The OECD has undertaken significant work in the area of digitalisation to understand and address the benefits, risks and potential policy responses for protecting and supporting financial consumers. The OECD has done this via its leading global policy work on financial education and financial consumer protection. In theory, using AI in smart contracts could further enhance their automation, by increasing their autonomy and allowing the underlying code to be dynamically adjusted according to market conditions.