In Systems

What Is Artificial Intelligence?

by Peter Purcell

Imagine sitting at the doctor’s office when a robot walks in, asks a series of questions, and then says, “The doctor will be with you in a moment.”

The doctor finally arrives with another robot trailing beside him. The doctor asks questions and performs a few tests, while inputting symptoms and details into his iPad. The robot actively processes all the information and disappears with the doctor to obtain “the results.” Upon return, the doctor provides a detailed packet that lists several health concerns and numerous home remedies.

What happened behind the scenes? What was the robot’s role? While the doctor had formed an initial hypothesis, the robot was able to confirm his thoughts and list all possible treatment options. How did it do that?

The robot was fed thousands of medical textbooks, countless research articles, and millions of web searches. As the doctor was asking questions, the robot was able to process all of this information almost instantaneously to come up with a diagnosis and recommendation equal to or possibly better than what the doctor could have provided.

Welcome to the world of artificial intelligence.

AI uses computers or machines to mimic human reasoning, learning, thought, and behavior. The backbone of AI is the concept of machine learning. Machine learning works by using algorithms (step-by-step mathematical rules) that tell the machine or computer how to behave. These algorithms are built so the machine or computer can understand concepts beyond what was programmed.

In essence, the algorithm sets the initial foundation of knowledge and the machine is able to take that knowledge and build upon it.

For example, an advanced AI machine was recently programmed to understand the basics of human behavior. Initially, it was taught simple things such as if someone leans in to someone else with their eyes closed and lips puckered, it usually indicates a kiss is about to happen. But, after watching hours of popular TV, that same machine now can predict more sophisticated human action and emotion before it occurs, according to an article in Popular Science magazine.

In this light, researchers studying AI often debate the definition and stages of the development of human cognition. At what point does a machine’s thinking become human-like? According to an article published by a computer science professor at Michigan State University, there are four sequential levels of AI:

Today’s AI capabilities put us in type two of the table above: Limited Memory AI.

This type of AI can track patterns in user behavior and use those patterns to predict basic future behavior. An example is smart or self-teaching thermostats. Smart thermostats track their users’ patterns of air use and learn to adjust the temperature depending on time of day and typical usage. Shifting toward type three, we have IBM’s smart computer, Watson. After winning Jeopardy in 2011, Watson moved on to more sophisticated applications, such as diagnosing diseases that were once impossible to understand.

Does this mean medical jobs are at risk? How about for other industries?

Author Dennis Gunton sparked a heated dispute when he stated, “Anyone who can be replaced by a machine deserves to be.” Harsh, right?

Indeed, many critics of AI are such because of the fear robots will take jobs away from human beings. And while that has been the case for industries such as manufacturing and call centers, automation will also create jobs in the long run. Although they are often more efficient, machines still require humans to develop, operate, and maintain them. At least for now, machines cannot accurately, genuinely emulate complex human qualities like empathy, critical thinking, and emotional intelligence. Regardless of anyone’s feelings about AI, one thing is quite clear: technology will continue to learn, adapt, and advance.

This article has been adapted from a chapter from Trenegy’s book, Jar(gone).



What to Know Before Implementing AI in Your Organization

by Todd Boutte

There’s a growing curiosity around the role of AI in business. People are wondering where AI fits into their organization, which AI apps are useful, and which can be ignored. ChatGPT set the recent AI surge in motion, making AI more tangible and accessible to the masses. It’s no longer a vague concept. We’re seeing real world use cases of AI influencing the way we work. However, as organizations shift their attention toward AI, there’s a lot of uncertainty around how to approach it.

It’s important to understand the implications of AI and how to think strategically about where it fits in your organization.

Below are a few recommendations for how to approach AI amid the hype and countless applications available.

1. Identify Where to Use AI

While AI holds great promise, not every application will be beneficial or necessary for your organization. It’s crucial to identify the areas where AI can provide real value to your business. This begins by identifying the challenges and pain points within your organization that could be alleviated with AI. Start by focusing on the problems that need solving rather than the solutions.

Some key questions to answer during this process:

  • Customer responsiveness – Are there areas in the organization slowing down customer responsiveness?
  • Repetitive tasks – Are there pockets of highly repetitive tasks requiring a lot of people to accomplish simple objectives?
  • Market competition – Where are the pain points when going head-to-head in the market? Where can you leapfrog the competition?
  • Processes – Are there bottlenecks in processes that prevent the organization from being nimble?
  • Knowledge sharing – Where are there opportunities to improve knowledge sharing within the organization?

Bottom line: Don’t implement AI for the sake of AI.

2. Understand AI’s Current Capabilities & Limitations

As of 2023, AI has strong capabilities, but it’s not a magic wand. AI is not capable of creative thinking, understanding human emotion, or strategic planning—these are areas where humans excel. AI can’t process nuance to the same degree. For now, AI should not be seen as a replacement for human labor, but as a powerful tool to make us more efficient.

To effectively implement AI, certain skillsets will be required to monitor, maintain, and continually reevaluate AI as it grows. So humans are still part of the equation.

3. Start Small and Scale Up

A common mistake organizations make when implementing AI is trying to do too much too soon. A better approach is to start small, test, and learn. Choose a specific process, task, or business function that could benefit from AI. Implement, test, measure the results, and learn from the experience. For example, you might use AI to analyze your sales data and identify patterns that can inform your sales strategy. Once you’ve seen success on a smaller scale, you can gradually scale up and apply AI to more complex tasks and larger business functions.

4. Develop an AI Strategy

Just like any technology or ERP implementation, incorporating AI requires thorough planning, analysis, and training. While many employees are using ChatGPT (and possibly a few other tools) for their own purposes, anything implemented across the entire organization will impact processes, roles, budgets, and overall operations.

Some of the major steps an AI strategy should include are:

Aligning the organization – A big hurdle during any implementation is aligning the organization around the initiative. Everyone needs to understand why the AI tool is beneficial, how it will be used, who has ownership, etc. Change can be difficult, so creating support early on is important. Organizations will also need an implementation team with the right skills, decision-making authority, and follow-through.

Understanding business requirements – What are current processes and how will they be improved with AI? For implementation success, business process requirements must be laid out step by step. Specifics are important when mapping out processes. It eliminates confusion and lack of direction in the long run.

Training and clear communication – Decisions should be communicated among the project team throughout the project, and when it’s time to roll out the new tool, training will help employees confidently use the new system. Training should provide employees with the initials tools and knowledge they need to understand the system. Usability directly influences adoption rates.

The strategy will likely also include selecting the right tool, finding an implementation partner, rolling it out, and refining roles and responsibilities—the same things required for a successful technology implementation of any kind.



Crucial Components of an AI Implementation

by Peter Purcell

There are some critical components when implementing AI in your organization. The general process isn’t all that different than implementing an ERP system or other type of software. It’s important to develop a strategy to obtain the maximum value out of this transformative technology. Below are some of the critical components to account for in your AI implementation plan. 

Define Objectives & Understand Business Requirements 

Determine which problems the AI tool will solve and how it will solve them. Don’t just implement AI for the sake of AI. It’s important to define where and how AI will provide real value to your organization. This involves mapping out current and desired future-state processes to find inefficiencies and understand where AI fits in. 

Turn general ideas into specifics. In doing so, you will eliminate confusion and establish a clearer vision. With a clear vision, it’s easier to create buy-in.  

Align the Organization 

Everyone should understand why AI is being implemented, why it’s beneficial, and how it will create efficiency. Change of any kind can be challenging, so it’s important to garner support early on.  

Aligning the organization also includes identifying who is responsible for each process impacted by AI and who will be responsible for maintaining the new technology going forward.  

Additionally, an implementation team will be vital for project success. Involve the people who have the right skills, will champion the new technology, and are committed to seeing the implementation through to completion. 

Select the Right Tool 

If several options are on the table, here are a few steps to narrow down your list.  

  1. Create requirements based on the business objectives for the new AI tool. 
  2. Leverage the requirements to perform research to identify and short list companies that provide tools supporting the requirements. Hint—use ChatGPT to help you here.  
  3. Request demos and ask for the companies to provide a clear understanding of functionality, cost, support model, and product development cycles.  
  4. Request references because AI is still relatively new, so certain products might be readily available but still in the testing phase. 
  5. Involve the right people who have valuable insight. Expanding involvement from team members can often plug holes and confirm whether or not the new tool will support the organization’s requirements and eliminate inefficiency.  
  6. Evaluate long-term value. Consider if and how the tool will be used in the next year and beyond. Is it scalable? Are the expected benefits worth the investment? 

Find an Implementation Partner 

It’s important to consider teaming with an implementation partner when adopting any new AI solution. Many companies have been in the AI space for a long time and know the ins and outs AI and how it functions in a variety of business environments. They are going to have in-depth knowledge on how to introduce such a powerful tool. They will know how to set boundaries, prioritize security, create buy in, and optimize your investment for the future. 

Rollout & Training 

Develop a plan for how you’ll roll out the new tool across the organization. When it’s time for rollout, provide role-specific, hands-on training to employees. Focus their training on how the AI relates to their role and how they’ll use it. Throughout training, encourage employees to share their feedback. Their insights can help identify areas that need additional focus or improvement. 

Establish Governance 

AI requires ongoing governance to monitor and maintain. It will need to be examined for accuracy on a regular basis to ensure it delivers correct, up-to-date information. It will also need to be examined for effectiveness and consistency to ensure that AI stays intelligent and isn’t misusing or misunderstanding information. We recommend establishing a regular schedule for maintenance and/or review. 



AI-Based Cyberattacks Are Here

by Peter Purcell

Hackers are creative, especially those backed by nation states. Does this mean companies should revisit their cybersecurity programs and invest in AI-based responses? AI-based cyberattacks sound ominous yet should not be treated different than any other attack. AI-based cyberattacks aim to gain access to information to steal money or gain access to control systems to cause harm. No different than today. The AI-based cyberattacks will simply make hacking faster. After all, AI systems don’t have to stop to drink Red Bull, go to the bathroom, or nap.

Companies can be well protected from AI-based cyberattacks with the following:

1. Operational systems

SCADA, ECM Monitoring, Power Grid and other process control systems often fall outside the purview of the IT organization. Basic table stakes security is often overlooked. For example, changing default passwords or updating software and firmware on a regular basis is a must. Software updates and passwords are vulnerabilities all hackers exploit when trying to shut down power grids or take over process control systems.

Functions responsible for operational systems should work with Internal Audit and IT to do the following:

  • Establish a cybersecurity champion (read more here)
  • Implement a cybersafety program
  • Change all default passwords and update all systems on a regular basis
  • Create and test recovery procedures (read more here)

2. ERP

Most ERP systems have been secured, but hackers will try to access employee information for identify theft or access vendor tables to change bank information to redirect payments. Companies should assume ERP systems are vulnerable and should establish the following basic controls:

  • Encrypt employee HR information
  • Strengthen employee awareness of possible hacking
  • Establish controls around the payment process
  • Monitor changes made to supplier bank information

3. Employee social media

Most employees have no idea what hackers can do with information posted on social media. Hackers spend significant time culling through social media to create and execute successful phishing and “spear phishing” campaigns. AI-based hacking can easily marry company website information with posts of pictures taken in the same country where a deal is being negotiated to acquire assets. Why? To create an email from the CEO requesting wire transfer of funds for a down payment. And the employee could fall for it.

Prevention is easy:

  • Create and communicate a social media vulnerability awareness program for all employees
  • Hold employees accountable for business controls compliance
  • Perform regular cybersecurity training

Companies have no choice but to live with the threat of cyberattacks. Hackers are relentless and creative. They will develop new and varied methods to steal money or cause problems by taking over any system. Unfortunately, companies fall into the trap of spending more and more money on unnecessary technology for protection. Compliance to common-sense controls is the best defense against hackers, whether human or AI-based.



by Todd Boutte

Hundreds of articles have been written debunking AI myths, and they all seem to say the same things. Below are some lesser-talked-about AI myths related to business that you might find helpful to consider. If AI is not already part of your business strategy, it likely will be in the next several years. It’s important to understand what’s true about AI and what’s not to make informed business decisions.

AI Myths

1. AI Can Work With Any Type of Data: It’s often assumed that AI can work effectively with any data type, regardless of its quality or format. However, AI models require well-organized, high-quality data to function accurately. Poor data quality can lead to ineffective AI solutions and inaccurate outcomes.

2. AI Systems Are Always Complex and Require Advanced Expertise: There’s a belief that AI systems are inherently complex and can only be managed by people with advanced experience in that field. While some AI applications are indeed complex, many user-friendly AI tools and platforms are designed for non-experts, making AI accessible to a broader range of users.

3. AI Implementation Always Leads to Quick Cost Savings: Many businesses assume that implementing AI will immediately reduce costs. While AI has the potential for cost savings, especially through automation and efficiency improvements, the initial investment, continuous development, and maintenance of AI systems can be substantial. The return on investment might occur over a longer period.

4. AI Guarantees a Competitive Advantage: While AI can provide significant benefits, merely implementing AI doesn’t automatically lead to a competitive edge. The strategic use of AI, aligned with a company’s business requirements and integrated effectively with its operations, is what can create a competitive advantage. Merely having AI technology is not enough.

5. All AI is Similar: People often think of AI as a monolithic technology, but AI encompasses a wide range of technologies and applications, from simple automation tools to advanced machine learning. Capabilities vary greatly and tools are specialized to help in many different areas and industries. We’ve seen AI in manufacturing, finance, the energy sector, human resources, accounting, IT, education, the auto industry, the U.S. military, and pretty much anything else you can think of.

6. AI Eliminates Human Error: AI systems are not immune to errors. They don’t magically produce the right answer every time. There might be programming flaws, biases in training data, or algorithm limitations. It’s important to review what information AI delivers, especially when dealing with financial data or legal matters.

7. AI’s Impact is Immediately Measurable: Some expect immediate, visible, and easily quantifiable impacts from AI. However, the benefits of AI might not be immediately measurable. There might be small wins as day-to-day tasks become automated or more efficient. But improved processes or efficiencies may become more significant over time. When implementing, give it time.

8. AI Ensures Business Success: Implementing AI is often viewed as a guarantee of business success. While AI can provide significant benefits, its effectiveness depends on strategic alignment with business goals, quality of implementation, and how well it’s integrated into broader business processes. Implementing AI for the sake of AI won’t make an organization successful.




Why Your Controller Should Influence Your AI Strategy

by Todd Boutte

Now that AI is an inseparable part of the workforce, it’s important for controllers to understand the impact of AI on their organization. AI is sneaking into organizations at an increasing rate. Employees are using ChatGPT, Bing Chat, Microsoft Copilot, and more to streamline personal workflow while nearly every software is releasing AI-powered features.

As a controller, there’s an inherent duty to ensure AI is utilized ethically, accurately, and in line with company objectives. The following is a guide for doing just that:

How Will AI Help Finance & Accounting?

AI has great potential—it just needs to be harnessed in the right ways. AI is already helping organizations in some of the ways listed below. With AI, Finance and Accounting organizations are saving time and money, freeing up employees to spend more time on activities that add value.

  • Efficiency & Automation: AI can automate routine accounting tasks, reducing manual efforts and errors, which leads to more efficient processes.
  • Data Analysis: AI-enhanced data analytics capabilities will help controllers derive insights from vast amounts of financial data, detect patterns, and make more informed decisions.
  • Fraud Detection: AI-driven systems can identify anomalies in financial transactions, helping in early detection of fraud or discrepancies that might go unnoticed in manual reviews.
  • Forecasting: AI can enhance the accuracy of financial forecasts by analyzing multiple variables and market conditions, aiding in better budgeting and planning.
  • Asset Management: AI can optimize asset utilization and monitor asset performance, crucial for protecting a company’s resources.
  • Competitive Advantage: Adopting AI in financial functions can give companies an edge over competitors in terms of speed, accuracy, and strategic decision-making.
  • Regulatory Compliance: AI tools can help ensure financial operations comply with ever-evolving regulations, reducing risks of non-compliance penalties.
  • Cost Savings: By automating processes and reducing errors, AI can lead to cost savings in the long run.

Awesome, but… 

There are still issues with AI that prevent controllers from fully relying on it. Controllers must understand the potential pitfalls and how they might affect the business in order to stay compliant.

  • Bias Issues: If data used to train AI models contains biases, the AI system can perpetuate or amplify these biases.
  • Loss of Privacy: Information shared with most AI tools ends up in the public domain.
  • Job Displacement: Automation through AI could render certain jobs obsolete.
  • Security Concerns: AI systems can be used to create malicious deepfakes that are difficult to detect.
  • Ethical Concerns: AI may be used to make recommendations or decisions in sensitive areas like healthcare or criminal justice.
  • Dependency: Over-reliance on AI systems can reduce human skill levels and lead to an inability to perform tasks without technological assistance.
  • Transparency & Accountability Issues: Many advanced AI models, especially deep learning models, act as “black boxes,” making it hard to understand how they arrive at specific decisions, which complicates accountability.
  • Safety Concerns: In areas like autonomous driving or robotics, AI malfunctions could have direct safety implications.
  • Existential Risks: A superintelligent AI could act in ways that are harmful to humanity if not properly aligned with human values.

It’s All About Protecting the Company

AI is powerful and can be useful, but it’s not infallible. The information and answers it provides are not always 100% accurate. For the controller, this is especially important when dealing with sensitive information, regulations, and finances in general.

It’s the controller’s responsibility to implement and monitor internal controls to ensure the integrity of an organization’s financial transactions and reporting, safeguard reporting, protect company assets, and ensure compliance. The controller should work with IT and Business to accomplish the following:

  1. Understand AI Basics: Gain a foundational grasp of AI’s capabilities and limitations.
  2. Set Clear Objectives: Determine the specific goals and intended outcomes of AI integration.
  3. Establish Governance: Create a governance framework with policies and guidelines for AI usage.
  4. Perform Risk Assessments: Evaluate potential pitfalls, inaccuracies, or vulnerabilities introduced by AI.
  5. Ensure Data Integrity: Prioritize the quality of data used to train and inform AI models.
  6. Monitor Compliance: Stay updated with relevant regulations and ensure AI systems are compliant.
  7. Provide Training: Equip teams with knowledge and skills to effectively use and manage AI tools.
  8. Implement Regular Audits: Set up periodic checks on AI performance, outcomes, and ethical considerations.
  9. Maintain Collaboration: Foster a strong relationship with IT and other departments for secure and effective AI deployment.
  10. Plan for Contingencies: Have backup processes in case AI systems malfunction or produce unintended results.

Bottom Line

AI is here and it’s not going away. Leveraging AI’s capabilities can drive innovation, optimize operations, and create competitive advantages—but discretion is crucial. Controllers have a responsibility to oversee how AI is used. By responsibly integrating AI, controllers can transcend traditional roles, becoming strategic executives guiding the company’s future.



The Future of Business Intelligence

by Todd Boutte

Business intelligence (BI) is a simple concept. It involves 1) collecting data pertaining to your company from internal and external sources and 2) finding a way to distill it into something actionable. Essentially it involves harvesting the data you need to make good business decisions.

Today, the term “business intelligence” usually refers to the software or tools organizations use to turn data into usable information. It’s come a long way in the last 10 years, and with the recent growth of artificial intelligence, BI tools have powerful potential.

How Business Intelligence Has Changed in the Last 10 Years

In the last decade, companies have consolidated their business systems around a few key players (Microsoft, Oracle, SAP). Microsoft stands out because they’ve built a product into an ecosystem that thousands of organizations use every day. Microsoft 365 developed Power BI as a stand-alone product within the last 10 years.

More than 10 years ago, tools were difficult to use. They required people with specialized skill sets to write code and gather data to turn it into usable information. Microsoft, however, as a leading data architect and office productivity company, has key contributions in the simplification of business intelligence. With Microsoft’s Power BI, companies no longer need an army of database administrators and developers to handle data. It has become more of an intuitive, self-service business intelligence platform people can use themselves.

The Future of Business Intelligence Is Artificial Intelligence

Since Microsoft is a key player in this industry, they’ve already included some basic AI tools within Power BI. One of those is a Q&A box that can be included in a report to make it easier to find information. A user can ask, “What was the revenue in Q1 of 2022 vs. 2023?” Power BI will do its best to pull that information.

As Microsoft continues to develop AI capabilities, we expect users will be able to ask even more complex questions and follow-up questions, just like with ChatGPT.

We also expect AI to be able to examine a set of data and make inferences based on that data. For example, suppose a company needs to review customer feedback on a product line but has 10,000 customer reviews. That’s a lot for one person to parse through. Instead, there’s potential for AI to step in and find common threads with language without having to spend hundreds of human hours looking through reviews. Instead of merely taking a 5-star review at face value, AI could analyze what was said about the product—because a 5-star review isn’t always meaningful if the text says otherwise.

It’s important to note that, no matter how advanced AI is, organizations shouldn’t fully rely on AI to extract data and make inferences. Human intelligence will still be required to manage AI tools and make sure they’re pulling the right data, making accurate inferences, and interpreting language correctly.

AI is all about saving time and allowing employees to add more value to the organization. As AI advances, it will alleviate a lot of time-consuming activity and allow employees to focus on the strategies and conversations that will drive business decisions.

The Right Mindset for AI

Remember, AI is a tool. It’s not a decision maker or a business strategist. While it can replace a lot of human tasks, it doesn’t replace a human. The people using AI tools are the key to making AI tools successful. The right tool in the wrong hands won’t solve anything. But if used correctly, AI has the potential to add significant value to organizations.

A Word on Best Practices

When it comes to managing data the following practices are crucial, with or without AI.

1. Establish Good Governance Around Data

Setting standards around creating data is key. We’ve seen companies that have multiple people entering the same data in their system under different names (e.g. GE, GE Power, General Electric). When someone asks to see information for GE, the data isn’t accurate. Organizations must have good governance and data ownership on the front end so information is centralized.

2. Define & Agree on Metrics

It’s important to agree on and define which metrics are to be tracked. Know what’s included in each metric and what’s not. If people realize data metrics aren’t consistent or correct, they won’t believe the data. They’ll be more likely to create their own databases that are more consistent with the data they need.

Terminology is critical. In many organizations, different departments or divisions within the company have different definitions for the same word. But there shouldn’t be any ambiguity on a well-built report. It’s important to note that BI software or tools can’t solve this problem. It’s about the processes around business intelligence and the people involved. BI software is maintained by people, and the processes for maintaining it must be clear and thorough.



ChatGPT: Common Questions & Potential Impact

by Todd Boutte

No consumer application has grown quite as rapidly as ChatGPT. If you aren’t familiar with ChatGPT, it’s an artificial intelligence chatbot that’s far more advanced than the average customer service chatbot. It was created by OpenAI using natural language processing to imitate a human as best as possible. The outcome is pretty impressive.

The AI chatbot reached 100 million monthly users in the first two months since its launch in November 2022. Even the largest and most profitable apps (Google, Facebook, YouTube) weren’t adopted at such a high rate.

It interacts in a conversational, humanlike way and has the ability to write and debug code, solve math equations, write full essays and articles, answer questions and follow-up questions, give instructions, and much more.

Common Questions About ChatGPT

The app has garnered some strong reactions—excitement, fear, skepticism. Will ChatGPT replace Google? Will ChatGPT replace employees? Can you trick ChatGPT into learning incorrect information? Some aspects of the app remain a little mysterious. Below, we address a few common questions people are asking.

How does it work?

While the exact ins and outs aren’t published, we asked ChatGPT itself where it gets its information. Here’s what it said: “ChatGPT has been trained on a diverse range of internet text to generate human-like responses to questions and prompts. This includes a wide variety of topics, such as news articles, scientific papers, historical documents, and fiction, among others. The model’s training data is sourced from the web, and its training process uses deep learning techniques to learn patterns in the text and generate responses based on that knowledge.”

It’s answer is a bit vague, but we know it’s doing more than just pulling responses from Google. It’s trained on a variety of sources, and it’s also continually learning from interactions with users.

Will ChatGPT replace Google search?

The app is currently intended to interact with people and learn, not serve as a search platform, although it does have similar capabilities. We don’t suspect it will completely replace Google search, at least not any time soon. However, Google will inevitably lose some traffic to ChatGPT as people figure out what it can do.

Think of it this way: Suppose you ask both Google and ChatGPT, “What is lease accounting?” Google will give you a list of sources on where to find that information. You’ll click the source that seems reputable and offers a digestible explanation. ChatGPT will give you one understandable explanation and answer follow-up questions. What ChatGPT doesn’t currently do is provide the most up-to-date information or offer insight from multiple sources, so it’s not a true replacement for Google.

In fact, the real competition against ChatGPT isn’t Google. It seems to be specialty sites and forums, some of which prohibit the use of ChatGPT altogether. These are places where people can get the back-and-forth interaction needed to solve problems and have actual conversations with people who know what they’re talking about.

Can you trick ChatGPT into learning incorrect information?

There’s not a solid answer for this, but if we had to guess, there’s probably not enough momentum to steer it. It would likely require millions of interactions. It’s not like trying to influence one person—it’s more like trying to influence hundreds of thousands of people at a time. ChatGPT gets it’s information from a large variety of sources, so to completely misdirect it would be difficult.

Will ChatGPT replace employees?

Most jobs require some level of human intervention, so it’s not likely to replace jobs. After all, ChatGPT has to have an input.

It does supplement and make some jobs easier. Take microblogging, for instance. Marketers and writers are already using ChatGPT to write or draft blogs and articles that require minimal edits. The job still requires a human to maintain and distribute articles.

ChatGPT also has the potential to make programming and developer jobs easier by writing complete source code. What it can’t do is peer review code. Developers can use the app to augment their work and remove some of the frustration and repetition. But anything ChatGPT produces still requires validation.

In short, it won’t entirely replace employees. It will just help them be more efficient and solve problems quickly.

The Future of ChatGPT

There’s a lot of room to grow with ChatGPT. Right now, we would classify it as a supplement—not a replacement—for your job or organization. Like any new tool, there are lots of possibilities and pitfalls, but it will likely become part of every organization’s technology arsenal in the next couple of decades.

For an even deeper dive into ChatGPT, listen to our podcast episode, ChatGPT: Implications for Business & Beyond, featuring Trenegy’s Technology Lead, Todd Boutte.



Considerations Before Using ChatGPT for Business

by Todd Boutte

With the rise of ChatGPT, we’ve heard a lot of talk around its potential in the business world. People are wondering how to use ChatGPT for business. We expect the app to have an influence across organizations as it continues to grow. Whether it’s ChatGPT specifically or another AI of ChatGPT’s caliber, companies should consider what changes might arise in the years to come. It won’t completely replace human jobs anytime soon, but it certainly has the potential to make organizations more efficient and effective.

For organizations that might be taking advantage of this technology in years to come, here are a few key considerations:

Identify Use Cases

Before any new technology implementation, it’s important to know exactly how it will be used and by whom. One major area in which we see ChatGPT (or a ChatGPT-like technology) functioning is knowledge management. Employees are almost always relying on tracking down a human, a manual, or a vendor when searching for information. What if their organization’s knowledge base was more portable and accessible? A virtual assistant with the conversational ability and accuracy of ChatGPT could potentially make knowledge management significantly more efficient. It would be like talking to a person who knows the ins and outs of every legal document, land record, contract, employee handbook, etc. who can answer follow up questions, make connections for you, and spot trends. But instead of a person, AI of this caliber has infinite bandwidth.

Find an Implementation Partner

It’s important to team with an implementation partner when implementing any new AI solution. There are many companies that have been in the AI space for a long time and know the ins and outs of a tool like ChatGPT. They are going to have the most knowledge when it comes to bringing such a powerful product into organizations. They will know how to set boundaries, prioritize security, and create buy in.

Most importantly, an implementation partner will help the tool pull information from the correct data by plugging in policies, procedures, and the entire framework from which the AI learns. Essentially, they will set the AI up to provide correct and complete information, including HR documents, instructions from vendors, legal documents, employee guidelines, troubleshooting tips, contracts, records, safety policies, and more.

Allganize is a natural language search solution company that has launched this type of solution. Check it out here.

Set up a Governance Model

Once ChatGPT or another similar tool is in place, it will require governance. It’s not simply a tool you set and forget. Organizations will have to treat it as both technology and “employee.” It’s almost like a new business analyst that requires more attention and training up front and eventually learns the ropes. As time goes on, organizations will still have to examine it for accuracy and completeness. But instead of requiring programming to improve its behavior, it relies on feedback.

The Key Takeaway

ChatGPT is a powerful tool that has the potential to save a lot of money and time. But it’s important to determine how you’ll use ChatGPT for business, partner with the right people, and ensure it’s managed properly. Right now, it seems most organizations are still in the consideration stage as they learn how it works and start to evaluate where it might best serve their business. Evaluating use cases is key to ensuring this type of tool will create value, efficiency, and a worthwhile return on investment for your organization.



by Peter Purcell

A recent article from Fortune reported a shortage of around 340,000 accountants in the US as of March 2024. Many large corporations’ recent quarterly earnings statements contained errors, and while we don’t really know if a shortage of accountants is to blame, it’s certainly a factor to consider.  

A lack of accounting staff may lead to overwork, which means a potentially greater margin for error due to longer working hours and having to balance multiple roles at once. Down the road, this could mean higher turnover rates. 

A shortage of accounting staff also means more job openings—but research shows that fewer and fewer college students have been majoring in accounting over the past several years. This could be attributed to many things: increased tuition, student loans for the extra year of school required to become a CPA, declining interest in accounting, the lack of perceived social impact, etc. 

Regardless, organizations need accurate and effective accounting processes to retain stakeholders and employees, increase revenue, meet customer needs, and generally operate as a successful company.  

By leveraging AI, the accounting profession can address major issues, like financial statement errors, the accountant shortage, the declining interest in accounting careers among younger generations, and the challenges associated with the education/certification process. AI has the potential to enhance the accuracy and efficiency of accounting practices and make the profession more attractive and accessible to a broader pool of talent. 

Mitigating the Accountant Shortage 

How AI can help organizations address the challenges associated with the CPA shortage:

1. Automating tasks – AI can automate routine and time-consuming tasks to alleviate workloads, allowing accountants to focus on what cannot yet be done with AI. This shift can help mitigate the effects of an accountant shortage by giving existing staff their time back.

2. Strategic hiring and compensation – Companies may need to reconsider their hiring strategies and compensation packages to attract talent in an AI-augmented accounting landscape. This could include offering roles that leverage AI skills, providing ongoing training in new technologies, and reevaluating the need for traditional credentials in favor of demonstrated competencies in AI and accounting software. 

3. Overall strategy – Consider how AI can be used to parse the organization’s financial trends and economic forecasts to identify where strategic challenges should be assessed. Determine how AI can help finance align the overall strategic plan with financial objectives, corporate structure, working capital, and funding. 

Reducing Financial Errors 

How AI can help identify and reduce accounting errors in financial reporting:

1. Enhancing accuracy and reducing errors – By automating routine tasks, AI can help reduce error or double check work to create more accurate financial reports. AI algorithms can help identify inconsistencies or deviations from expected financial norms. 

2. Predictive analysis – AI can provide insight into trends by analyzing large amounts of data quickly. Predictions can help accountants focus review efforts where errors are most likely to occur. They can also help with financial planning and risk management. 

Addressing Educational & Entry Barriers 

How AI can help address education and certification challenges associated with the accounting profession:

1. Revolutionizing accounting education – AI tools can provide more interactive and personalized learning experiences that incorporate real-world scenarios. There’s potential for students to gain hands-on experience with AI to bridge the practical experience gap. Perhaps this could integrate more practical experience into the standard education timeline and alleviate an accountant shortage in the long-run. 

2. Continuous learning and adaptation – Both students and professionals in the accounting field need to adopt a mindset of continuous learning to keep up with technological advancements. AI can help with deeper learning and staying up to date with accounting and controls practices.  

Attracting the Next Generation 

How organizations can attract and retain accounting talent:

1. Highlighting the role of tech in accounting – Emphasizing the integration of AI and other technologies into accounting roles can make the profession more appealing to tech-savvy individuals. Showcasing how AI can elevate the accountant’s role from repetitive tasks to more analytical and advisory functions can attract individuals interested in technology and innovation. 

2. Addressing education cost concerns – AI can also play a role in creating more efficient pathways to certification. For example, AI-driven platforms could offer alternative credentialing mechanisms or support programs, making the career path more accessible financially. 

As accounting adapts to incorporate AI, the role of accountants will evolve to hopefully emphasize more strategic, advisory, and analytical skills to attract new talent. However, implementing AI requires thorough planning and a robust strategy. We would love to share how we’ve helped finance organizations achieve success through automation and how we can help your organization do the same. Email us anytime at 



How to Manage M&A Clean Rooms with AI

by Todd Boutte

Clean rooms are especially common in merger and acquisition (M&A) transactions involving highly sensitive industries or when dealing with proprietary technology, trade secrets, or other valuable intellectual property. By establishing a clean room, the buyer and seller can engage in discussions and analysis while minimizing the risk of information leakage that could potentially jeopardize the deal. 

Introducing AI into a clean room environment is helpful as long as it’s secure and procedures are in place to prevent unauthorized access or disclosure.  

Microsoft SharePoint, Copilot, and Power Automate are great tools to support M&A efforts and keep teams organized and focused. They can assist in managing M&A clean rooms for teams in the following ways: 

Microsoft SharePoint 

Helpful for: Document and file management 

  • Document Management: SharePoint provides a robust platform for document management. In M&A clean rooms, you can create dedicated SharePoint sites or libraries to store confidential documents related to the transaction. 
  • Access Control: SharePoint allows fine-grained access control. You can restrict access to authorized team members so only relevant individuals can view or edit sensitive information. 
  • Versioning & Auditing: SharePoint tracks document versions and maintains an audit trail. This is crucial during M&A negotiations, as it ensures transparency and accountability. 
  • Collaboration: Teams can collaborate on documents within SharePoint, making it easier to review, annotate, and discuss critical files in real time. 

Microsoft 365 Copilot 

Helpful for: Efficiency and productivity 

  • AI-Powered Assistance: Microsoft 365 Copilot leverages large language models (LLMs) to assist users in various Microsoft 365 apps (Word, Excel, PowerPoint, etc.). It’s embedded in these apps to help users be more efficient. 
  • Creativity & Productivity: Copilot generates drafts, suggestions, and content based on natural language prompts. For example: 
  • In Word, it jump-starts the creative process by providing a first draft that you can edit and iterate on. 
  • In PowerPoint, it helps create presentations by adding relevant content from your existing documents. 
  • User Control: You remain in control as the author. You can modify, discard, or accept Copilot’s suggestions. 

Microsoft Power Automate (formerly Microsoft Flow) 

Helpful for: Automating processes and workflows 

  • Workflow Automation: Power Automate allows you to automate repetitive tasks and processes. In M&A clean rooms, users can set up flows to trigger actions based on certain events (e.g., document uploads, approvals, notifications). For instance, when a new document is added to a SharePoint library, Power Automate can notify relevant team members or initiate an approval workflow. 
  • Integration with Other Apps: Power Automate integrates seamlessly with various Microsoft 365 apps and third-party services. You can create custom workflows that connect SharePoint, Teams, Outlook, and more. 
  • Data Transformation: Power Automate can transform data between different formats to ensure consistency and compatibility across systems. 

Together, these apps contribute to efficient and secure management of M&A clean rooms, enabling teams to collaborate effectively while safeguarding sensitive information. 

Key Considerations When Using AI with Confidential Data 

When AI is involved with confidential information or sensitive data, teams must be aware of the potential risks. The following are best practices when using AI, and any technology for that matter. 

  1. Be purposeful with how you use AI. Make sure it contributes to your organization’s overall strategic objectives. 
  1. When using AI for company purposes, make sure you have approval before downloading or feeding information through any new AI software, especially when sensitive information is involved. 
  1. Avoid entering confidential, secure, or proprietary information into external AI tools. Instead, remove the confidential or identifying information first, and then run it through. 
  1. Don’t accept every AI output as truth. Sometimes AI delivers incorrect information or data, so verify the accuracy of anything AI produces.

For more info on implementing these AI tools, or for an overall more successful M&A transition, reach out to us at 



Where AI Fits into the Future of Finance

by Bill Aimone

Finance and accounting professionals are asking a variety of questions about how AI will impact the future of finance. If you’re reading this now, you’re likely in one of two positions:

  1. You’re charged with understanding how the entire finance organization should strategically benefit from AI to enable the organization to be more competitive.
  2. You want to know how you can use AI to improve your personal productivity, which we’ll discuss at the end of this article.

To be strategic, finance organizations should view AI from the perspective of Finance’s overall role in the company. Below are the key questions finance executives can start with:

1. Process

How can we optimize business processes to align Administration with Sales and Operations using AI?

Standardizing repetitive processes using AI is a great starting point for the finance organization. Take OCR (optical character recognition), for example. OCR has advanced to the point where AI is embedded in the software. This allows for smarter recognition of information on invoices, which makes it easier to capture info in the ERP. It makes for an easier process and saves time that would have been spent on a repetitive task.

2. Strategy

How will AI help Finance align the strategic plan with financial objectives, corporate structure, working capital, and funding?

Consider how AI can be used to sift through the company’s financial trends and economic forecasts to identify where strategic challenges might need to be assessed. An example of this is using AI to analyze assumptions in the business plan along with changing economic and geo-political environments to identify what needs to be reassessed. An even simpler example is having AI assess and identify updates in a SWOT analysis conducted 5 years ago against the changing business environment over the years.

3. Control

How will AI help Finance safeguard assets through efficient and effective risk management controls (policy, process transactions, compliance, regulatory, cash management)?

AI can be used to review internal data and extract information to pinpoint where current risks may need to be reassessed. It can also be used to generate updated policy documents based upon changes in the business environment. A simple example is having AI automatically generate FAQs in a new policy document to aid in communications.

4. Information

How can we use AI to provide quality information to stakeholders more quickly to support decision-making, goal setting, regulatory requirements, and planning?

AI tools are already used in business intelligence to identify trends and develop smart forecasting models based on a series of data lakes the organization develops with internal information. Consider how it can help speed up the communication process.

AI for Finance Toolkit

AI has applicability within each of the above and a deeper dive into the supporting process that finance and accounting supports as a part of a finance AI initiative will a lynch pin for advancing just about every organization.  Trenegy has developed an AI for Finance Toolkit that allows finance executives to quickly prioritize AI opportunities. Our AI for Finance Toolkit has been embedded into our finance transformation methodology along with leading practices and a comprehensive look at tools available in the market to jumpstart a finance AI initiative.

Using AI to Improve Personal Productivity

From a tactical perspective, any professional in the finance organization can leverage AI as part of their job. Our advice to finance and accounting professionals is to view AI as a productivity tool. Find ways to use AI to make things easier. For example, Microsoft Copilot has all sorts of mind-blowing capabilities that can advance and automate the finance and accounting process. Using Copilot is akin to learning the capabilities of Excel to automate reporting, allocations, and forecast models. It’s simply a tool designed to make your job easier.

Copilot can be used to mine information and create summaries and documentation using a combination of internal SharePoint files and external information. This could help with tasks like updating management disclosure documentation, identifying anomalies in financial data, and documenting budget meeting results. The opportunities are endless.

One Final Thought

Finance organizations can leverage AI to enhance their value proposition and become more agile, innovative, and collaborative. However, implementing AI in finance requires clear guiding principles, a robust roadmap, and disciplined execution. That’s why Trenegy offers a comprehensive “AI for Finance Toolkit” that helps finance leaders assess their current state, identify high-impact opportunities, select the best-fit solutions, and manage change effectively.

Our toolkit combines our deep expertise in finance processes, data analytics, and change management with our extensive knowledge of the AI landscape and leading practices. Whether you want to automate your reporting, forecasting, compliance, or planning functions, we can help you design and implement a tailored AI solution that meets your needs and goals.

Reach out to learn more about Trenegy’s “AI for Finance Toolkit” and how we have helped other finance organizations achieve success through automation. We would love to hear from you and explore how we can support your finance transformation journey. Email us anytime at

facebook linkedin twitter

Recent Posts

Start typing and press Enter to search