AI continues to disrupt global businesses in new and exciting ways, demonstrating its ability to streamline processes and reduce manual intervention, thus allowing professionals to focus on higher-priority business tasks. This is excellent news, especially for finance teams, which by their very data-driven nature are ripe for the integration of AI into their operational framework.

AI-powered tools are becoming increasingly sophisticated. Their ability to swiftly analyze large amounts of data and their affinity for rapid learning is now being deployed in complex functions such as risk-proofing organizations, making predictions about competitors, helping finance and accounting teams in budgeting and forecasting, and eliminating risks in regulatory compliance.

A survey by Gartner reveals that 78% of the participating CFOs plan to maintain or increase enterprise-wide digital investments in the next two years – despite planning cost reductions if inflation persists.¹

Lessons from the pandemic: recognizing the gaps

While the role of digitalization in finance has always been well-acknowledged, its adoption accelerated due to COVID-19. The pandemic compelled organizations to free their finance teams from paper dependencies and adopt cloud-based technologies and automation to operate efficiently from remote locations.

In a survey, finance leaders looked back at their teams’ performance during the COVID-induced lockdowns and reflected on the gaps they faced in terms of their ability to execute in response to the crisis (refer Chart 1: Challenges faced by finance teams during the pandemic).² The responses can be broadly categorized into three segments:

• Insights-readiness: Being decision-ready with accurate, timely data to help management assess, plan, and execute quickly in a rapidly changing environment.
• Automation-readiness: The ability to be automation-ready, with touchless transactions and capabilities that supported virtual work and digital commerce.
• Change-readiness: The need to be change-ready with the right people, technology, and analytical skill sets to manage emergencies such as the pandemic.

Chart 1: Challenges Faced by Finance Teams During the Pandemic

Dimension of Readiness	Percentage	Core Operational Focus
Being decision-ready with accurate, timely data to help management assess, plan, and execute quickly	45%	Insights-readiness
Being automation-ready with touchless transactions and capabilities that supported virtual work and digital commerce	39%	Automation-readiness
Being change-ready with the right people, technology, and analytical skill sets in place to manage through the pandemic	15%	Change-readiness

AI tools can help finance teams to bridge these gaps and make them even more future-ready.

The future of organizational finance will be driven by AI tools

“Finance automation has effectively eliminated the need for human intervention in repetitive tasks while ensuring process excellence.”

Finance automation has effectively eliminated the need for human intervention in repetitive tasks while ensuring process excellence. The Hackett Group’s 2021 Key Issues Study found that finance and accounting functions are the most automated, with 79% of the respondents reporting that they have implemented automation in these areas. High digitalization of the finance function results from the tangible benefits that organizations achieve - automating paper-based, error-prone financial processes and digitizing financial data provide efficiencies and better visibility to help future-proof the entire organization.

AI can further accelerate the quest for finance automation by providing a range of capabilities that enhance and complement existing solutions. The technology can assist finance teams in the smooth execution of tasks relating to several areas of digital transformation (refer Chart 2: Top digital transformation priorities of the finance function).³

Chart 2: Top Digital Transformation Priorities of the Finance Function

Priority Area	Share of CFO Priority
Data management & analysis	49%
Financial close, consolidation & external reporting	40%
Management reporting & analysis	38%
Audit & risk management	30%
Scenario planning, budgeting & forecasting	24%
Order to cash	10%
Source to pay	6%
Asset management	2%

Here’s how:

1. Automation

• By automating repetitive, manual tasks such as data entry and invoice processing through the integration of third-party software, AI can provide a strong foundation for seamless and efficient processing.
• AI can be used to build chatbots and virtual assistants to help finance teams automate tasks, answer common questions, and support team members by taking over mundane and repetitive tasks.

2. Analytics

• By accurately and swiftly analyzing a large amount of financial data and presenting it in an easy-to-understand format, AI assists finance teams with data insights for informed decision-making.
• By processing natural language data that is in an unstructured format (such as emails and feedback), AI can help finance teams gain insights into areas of the business that need improvement, identify trends, and make data-driven decisions.

3. Integration

• By deploying machine learning algorithms to analyze trends and patterns in historical data, AI tools can help finance teams make informed decisions about budgeting, forecasting, and risk-management.
• When integrated with finance systems such as Enterprise Resource Planning and accounting software, AI tools enhance their capabilities and improve their output efficiency.

4. Security

• AI tools can identify potential cyber security risks, by identifying fraud patterns, and recommend robust cyber security measures to protect financial data and prevent cyber-attacks.
• By identifying potential risks to the organization, such as fraud, compliance violations, or revenue leaks, AI tools can help finance teams mitigate risks and ensure compliance with regulatory requirements.

5. Customization

• Finance teams can customize AI tools for specific requirements such as developing custom models for financial forecasting or scenario planning.
• By providing instant messaging, cloud-based solutions, and project management tools, AI tools can facilitate seamless communication among team members regardless of location and access to critical data from anywhere.

6. Support in Decision-Making

“AI tools can help finance teams with the latest industry-related information to support them in decision-making.”

AI tools can help finance teams with the latest industry-related information to support them in decision-making. For instance, ChatGPT can answer questions about financial regulations, tax laws and investment opportunities.

However, the exact implementation and integration of AI platforms will depend on the specific requirements of the enterprise. Therefore, it is essential for finance leaders to carefully consider their goals and objectives and plan a comprehensive implementation strategy to ensure maximum benefits.

With their potential to contribute to a wide array of organizational finance responsibilities – from managing cash flow to mitigating compliance risks – AI can successfully elevate the role of finance to new heights.

AI technologies can bridge strategy and action

In a survey, finance leaders identified data management and analysis as a critical priority for their teams. The second step is to convert this data into actionable insights that will lead to a business advantage.⁴ The increased tech investments come in the wake of expanding responsibilities and strategic challenges that finance teams are increasingly confronted with. These range from supply-chain management to risk-mitigation to ESG.

Since the decisions of the finance department impact an enterprise as a whole, it needs to understand how an organization functions and its performance. It is, therefore, more important than ever for finance teams to deploy the right technologies.

AI tools can assist by streamlining mission-critical processes and reducing mundane, manual work. Additionally, they can provide a fresh perspective into data, which is especially valuable as organizations face an increasing array of strategic challenges in a volatile and uncertain economy.

How can organizations deploy AI models such as ChatGPT in their finance functions?

While identifying the platform is the first step, integrating it successfully into the organization’s finance function is the next. I asked ChatGPT to assist with this. Here’s the response I received:

1. Identify the business problems and use case: The first step is to identify the specific business problem that the organization wants to solve with ChatGPT. For example, it could be to improve data analysis, decision-making, or customer service. The use case will determine how ChatGPT will be trained and integrated into the organization's workflows.
2. Data preparation and training: The next step is to prepare the data that ChatGPT will use to learn and perform its tasks. This involves identifying the relevant data sources, cleaning and preparing the data, and labeling the data to train ChatGPT. The organization may also need to hire a data scientist or machine learning expert to train and fine-tune ChatGPT for the specific use case.
3. Integration with existing systems: Once ChatGPT is trained and ready, the organization will need to integrate it into its existing systems and workflows. This may involve working with software developers or IT teams to integrate ChatGPT with existing data analytics, decision-making, or customer service tools. The organization should also establish clear protocols for how ChatGPT will be used in conjunction with human decision-makers.
4. Testing and evaluation: After ChatGPT is integrated into the organization's workflows, it should be tested and evaluated to ensure that it is performing as expected. This involves monitoring ChatGPT's accuracy, speed, and effectiveness in addressing the business problem. The organization should also establish key performance indicators (KPIs) to track the success of the ChatGPT integration.
5. Maintenance and continuous improvement: Finally, the organization should plan for ongoing maintenance and continuous improvement of ChatGPT. This may involve monitoring ChatGPT's performance, retraining it as necessary, and staying up-to-date with the latest advances in natural language processing and machine learning.

Overall, integrating ChatGPT into the finance function requires a multidisciplinary approach that involves data science, software development, and business strategy. By carefully planning and executing each step, organizations can leverage the power of ChatGPT to enhance their finance function and achieve better business outcomes.

Demonstrating the usefulness of AI tools in finance functions

The increasing use of AI tools by leading financial services providers is testimony to its usefulness, scalability, and expediency in the finance function. Here are some examples:

• Improved customer experience: Customized, AI-powered chatbots can play an important role in the overall customer experience. For instance, chatbots can answer user queries by fetching relevant information from the company’s repository almost in real time. With the right customization, these chatbots can answer thousands of queries a day with a high degree of accuracy and participate in intelligent conversations, thus enhancing the customer experience.⁵
• Increased efficiency: Intelligent automation – a combination of Robotic Process Automation and AI – can reduce manual effort and improve operational efficiency. By saving thousands of hours of manual labour, they can also result in colossal savings, annually.⁶
• Improved risk-management: Companies suffer significant losses due to fraudulent transactions, legal fees, investigation, recovery expenses and other related factors.⁷ Now industries such as banking and financial services are using AI technologies to detect identity fraud and lending fraud effectively, thereby helping to risk-proof the organization. They are also being successfully deployed to reduce regulatory compliance risk.

Taking the AI leap costs time and money

“Despite its innumerable benefits, the cost of integrating AI models into financial services tends to make finance teams hesitate before taking the plunge.”

Despite its innumerable benefits, the cost of integrating AI models into financial services tends to make finance teams hesitate before taking the plunge. Studies reveal that many finance leaders believe in starting small to avoid expensive failures with technology investments. They are prepared to embrace widespread tech adaption in their day-to-day functioning only when they perceive concrete benefits.⁸

Integrating AI models into a business can be affordable or expensive, depending on many factors. According to industry benchmarks, in 2023, companies can pay from $0 to more than $300,000 for AI software. Their options range from off-the-shelf-solutions to custom-built platforms developed by a team of in-house or freelance data scientists.⁹ Typically, pre-built solutions are cheaper than customized ones. Further, the cost is also impacted by the function expected from the AI solution: eg. virtual assistance, chatbots, or analysis. The management of the platform – how an organization develops, launches, and manages it – also adds not just to the cost of implementing it but also to the responsibility of keeping it running smoothly. For instance, in-house management gives the organization complete control over the solution. But the costs associated with maintaining an internal team will impact the company’s bottom line. Alternatively, an outsourced management model may cost less but this exposes the organization to the risk of sharing their confidential data with third parties.

The cost also varies with the time taken to integrate AI into the company’s processes. Most AI transformations take 18 to 36 months to complete, with some taking as long as 5 years.¹⁰ The timeframe depends on the scope and complexity of use case.

AI integration: Assessing the qualitative and quantitative dimensions

While the investments in implementing AI technologies can be significant, the potential benefits can also be substantial. To make the most of the technology’s potential, finance leaders must carefully evaluate its costs and benefits and develop a comprehensive strategy for its integration into their workflows.

Quantitative Advantages	Qualitative Advantages
Cost savings through reduction in manual labour	Facilitates informed decision-making by providing real-time business insights
Error reduction by processing large amounts of data with greater accuracy	Improved stakeholder experience by reducing response time and providing personalized support
Reducing time and effort by automating and streamlining processes	Better risk-management and more innovation
Time savings by computing financial data at an accelerated pace	Competitive advantage by enabling users to identify market trends and respond more quickly to industry changes

To AI or not to AI: Challenges in AI implementation

Admittedly, AI tools come with risks. It is, therefore, critical that finance leaders review them in the context of the organizational situation and assess the benefits before integrating an AI solution into their processes. Some of the risks and their mitigation strategies are listed below:

1. Data quality: AI tools require large amounts of data to be trained properly. Often, business decision-makers underestimate the time it takes to do ‘data prep’ before a data science engineer or analyst can build an AI algorithm. This critical stage is the foundation for the entire project. Inaccurate or incomplete data can adversely impact the quality of the output.
2. Interpretability: Given the complex technology used to analyze patterns in the large amount of data fed to AI-led platforms, it is often challenging to comprehend the logic behind their decisions and predictions. This lack of interpretability makes it difficult to identify errors or biases in the model’s output. One way of countering this challenge is by developing techniques to visualize the model’s decision-making process to explain the model’s output in more accessible terms.
3. Cyber Security: AI tools in finance functions access large amounts of confidential and sensitive information. They are, hence, vulnerable to cyber risks such as data breaches, malicious attacks, or model poisoning – leading to financial losses for the organization. To counter these risks, finance leaders should take proactive steps. These include implementing strong access controls, encrypting sensitive data, monitoring for suspicious activity, and regularly updating software and systems to address known vulnerabilities. Additionally, organizations should have a robust incident-response plan in place to respond to potential cyber security incidents promptly.
4. Scalability: It is often seen that while pilot projects yield small gains for organizations, teams often struggle to scale them to the company level or integrate them with legacy systems. For AI models to address business problems, improve existing processes, and deliver concrete results over the long term, they should be based on multi-perspective analysis. Hence, before investing in these programs, the implementation team should conduct exhaustive research and a detailed analysis. This will enable them to make AI a worthwhile investment for the finance function.

AI cannot replace human ingenuity

A note of caution: While their potential is impressive, AI tools are not a replacement for human creativity and resourcefulness. This is because:

1. Limited context understanding: Financial decision-making involves numerical data analysis and non-numerical factors such as market trends, business strategy, and industry knowledge. While AI tools can simplify the process of calculation, organizations need to depend on their team members to make decisions based on context and experience.
2. Lack of empathy: Finance professionals deal with sensitive information, and while dealing with this, team members have to exercise judgment, creativity, and emotional intelligence. Also, financial organizations require trust and confidentiality, something that no technology, regardless of how advanced it is, can claim to possess fully, as yet.
3. Limitations in data quality: Another concern is the potential for bias in the algorithms used to analyze data, which could lead to unfair or discriminatory outcomes.

It is vital for finance leaders to recognize that AI models can supplement human expertise and only partially replace these attributes. Therefore, organizations should deploy AI tools in conjunction with human oversight and intervention.

Upskilling and reskilling are the currencies of AI-led enterprises

Despite making considerable investments in AI, many organizations are yet to report business gains from the technology. Companies must restructure their corporate frameworks and train their teams to utilize these technologies fully. In other words, any investment in technology necessitates an equal investment in human talent. Such a strategy is essential to empower employees to extract value from the data provided by advancing technologies. Once talent strategies and business goals align, organizations can utilize people data to identify critical roles and skills, and the areas that require reskilling to drive maximum value. Importantly, reskilling should be a continuous process to ensure that the workforce’s agility and competitiveness are at par with the technology deployed.

AI can lend a cutting-edge advantage to the future of finance

The finance function is under constant pressure to evolve, to address the needs of the dynamic business landscape - which itself is witnessing exciting changes owing to the rapid integration of digital technologies across functions. To thrive in this environment, forward-looking CFOs must think ahead of the curve to successfully create a finance function that will proactively add value to tomorrow’s enterprises. AI tools such as ChatGPT can provide vital support towards this goal. At the same time, organizations must be mindful of the limitations of such technologies. To balance the benefits of AI against its potential risks, CFOs and CTOs must work together to implement appropriate safeguards. In addition, a collaborative effort is imperative to ensure that the use of AI is responsible and ethical.

The CEO of a research company sums up the scenario aptly by stating, “We have an unprecedented, once-in-an-era opportunity to make rapid, fundamental changes to the way we design and run our businesses. This opportunity forces us to rethink our skillsets, our careers, and the places where we work. This is a time to revisit those values important to us and to challenge our appetite for learning new techniques and ways of conducting business.”

---

Footnotes & References

1. Gartner: https://www.gartner.com/en/articles/how-your-cfo-cio-partnership-drives-digital-funding-or-not
2. Workday CFO Indicator Survey: https://forms.workday.com/en-us/other/cfo-indicator-survey-report-infographic/form.html?step=step1_default
3. Workday Survey Infographic on Digital Priorities: https://forms.workday.com/en-us/other/cfo-indicator-survey-report-infographic/form.html?step=step1_default
4. Workday Strategic Survey: https://forms.workday.com/en-us/other/cfo-indicator-survey-report-infographic/form.html?step=step1_default
5. Revechat: https://www.revechat.com/blog/chatbot-examples/
6. Cognizant Softvision Case Study: https://www.cognizant.com/en_us/case-studies/documents/cognizant-softvision-saves-big-4-accounting-firm-8-million-annually-codex5198.pdf
7. RapidMiner: https://rapidminer.com/blog/3-ways-ai-transforming-risk-management-banking/
8. Gartner CFO Mindset Shifts: https://www.gartner.com/en/finance/trends/3-cfo-mindset-shifts-autonomous-finance
9. WebFX AI Pricing Benchmarks: https://www.webfx.com/martech/pricing/ai/
10. Harvard Business Review: https://hbr.org/2019/07/building-the-ai-powered-organization

---

Author may be reached at: manojkalra@rediffmail.com and eboard@icai.in