Finance AI: How Different Functions — Like Scammer Detection And Portfolio Management — Are Improved By AI

Here are some of the most popular ways AI is already being used in the finance industry.

Portfolio management is the act of an individual or company selecting and controlling a group of assets, like bonds, cash and stocks that meets their long-term financial goals. According to the CFA Institute, AI can be used to conduct "sophisticated" analysis and optimize asset allocations in portfolios. For example, AI can provide estimates of returns that can be used for optimization. AI can also be used to read and assess earnings transcripts, identify relationships between securities and market indicators, and monitor and analyze search engines for words on key topics to help form hedging strategies, according to a report from Deloitte. The report also claims Man Group, a hedge fund, has used AI to assist with investment management, and as a result, their assets have quintupled since 2014. Some companies that focus on using AI to help with investment are Kavout (clients include Crimson Black Capital, Benzinga and FactSet), Trade Ideas (clients include Interactive Brokers, NYSE and Nasdaq) and Sentieo (clients include Schroders, BMO Capital Markets and MFG Asset Management).

Many insurance companies are already using AI in multiple ways, like underwriting, claims processing, fraud detection and customer service. One of the most user-facing ways is through the use of customer service chatbots. Lemonade, Allstate and Geico all use chatbots to give basic advice on things like billing information and address common questions about transactions. A report from the National Association of Insurance Commissioners found out of the 193 insurance companies surveyed, 135 reported using AI or machine learning for claims operation. AI can help make claims handling more efficient by automatically reading, interpreting and processing documents and photos, categorizing incoming cases to drastically reduce the case handling time for customers and making plausible assessments for different types of injuries, according to a report by Insurance Europe. AI can also help insurance companies with competitive pricing, like using the technology to collect information about a person's marital status, geographical location and likelihood of filing a claim, and then setting premiums based on this information.

Many insurance companies are already using AI in multiple ways, like underwriting, claims processing, fraud detection and customer service. One of the most user-facing ways is through the use of customer service chatbots. Lemonade, Allstate and Geico all use chatbots to give basic advice on things like billing information and address common questions about transactions. A report from the National Association of Insurance Commissioners found out of the 193 insurance companies surveyed, 135 reported using AI or machine learning for claims operation. AI can help make claims handling more efficient by automatically reading, interpreting and processing documents and photos, categorizing incoming cases to drastically reduce the case handling time for customers and making plausible assessments for different types of injuries, according to a report by Insurance Europe. AI can also help insurance companies with competitive pricing, like using the technology to collect information about a person's marital status, geographical location and likelihood of filing a claim, and then setting premiums based on this information.

According to IBM, $44 billion is projected to be lost due to fraud by 2025 and 72% of business leaders list fraud as a growing concern. Companies use AI to assist with online fraud detection and prevention by training the software with their historical data so it can suggest risk rules. These rules can then be implemented to allow or block user interactions like suspicious login attempts, fraudulent transactions and identity theft. According to Fortune, top companies like Google, Amazon and Microsoft all use AI-powered fraud detection services to help identify hackers and suspicious logins. There are downsides, however, like the software's tendency to flag real people logging into accounts. Fortune reports a Microsoft system that protects customers from fake logins has a 2.8% rate of false positives. In banking, there are three common types of fraud: phishing scams, identity theft and unauthorized transactions. In 2021, Fintech News reported financial institutions have spent $217 billion on AI to help prevent fraud. Banks can use AI to develop fraud scores to determine legitimate and fraudulent transactions, create purchase profiles to get a sense for customer behavior and compile a list of suspected attempts that can then be investigated, informational technology company, Telus International, recommends.

According to GCC FinTax, AI can be used to increase tax collection, develop precise profiles of each customer, reduce tax evasion by detecting irregularities, make more efficient calculations and reduce the time spent on audits. The Forum on Tax Administration's 2022 report found over 70% of tax administrators surveyed use innovative techniques like AI to uncover previously hidden assets or identify new risks. A report by EY found tax teams use 40% to 70% of their time gathering and manipulating, something AI can do in "a fraction of the time." However, EY believes one of the biggest reasons why tax teams aren't implementing AI is a lack of trust. A senior manager for the company said tax professionals are expected to "stand by their decisions...and show the work that supports those decisions," noting AI should be held to the same standards. Companies like H&R Block, Intuit and FlyFin use AI in their tax processes.

Bard Vs. ChatGPT: The Major Difference Between The AI Chat Tools, Explained (Forbes)

Here's What To Know About OpenAI's ChatGPT -- What It's Disrupting And How To Use It (Forbes)