Artificial Intelligence (A.I.) is here. Not long ago considered the stuff of sci-fi movies and select industries, A.I. has moved into the mainstream. However, many companies have been slow to catch on and see how data can help with the risk analysis of critical business decisions. Pinkerton President Jack Zahran and Vice Chairman Tim Williams say that A.I. is forcing companies to rethink their attitudes and embrace data as a key element in their security operations.
Generating usable knowledge from noisy data
“There has been an explosive growth of applications that are generating immense amounts of data,” says Williams. “But so much of it is noise. The problem companies have is parsing this data into usable knowledge. That’s where A.I. comes in.”
Trying to find relevant data that can impact your business decisions is akin to trying to sip water from a firehose. There is just too much to consume. “Data collection is not enough,” says Zahran. “Having the sophistication and resources to analyze that data and turn it into actionable intelligence will make the biggest difference.”
The credit card industry has been on the forefront of using A.I. for many years. When an individual uses a credit card, information is collected. This information includes items purchased, month/day/time of the purchase, price, store location, salesperson completing the transaction. While much has been reported about the use of data for marketing, the credit card industry also applies A.I. to the data to enhance protection of themselves and their customers from losses.
As an example of the application of A.I., consider the data and technology that come together whenever a consumer gets an automated call alerting them to a potentially fraudulent purchase. For that to happen, the credit card company’s system has to pinpoint this transaction from millions of others, analyze it in comparison to the customer’s past purchase history, determine that one or several factors indicate possible fraud, trigger the automated call procedures and have a protocol in place if the consumer confirms that it was a fraudulent purchase. All of this takes place nearly instantly.
Using AI for decision making
With so much data available to companies, one might think that better decisions are being routinely made today compared to a decade or more ago. Williams doesn’t think so. “A lot of critical business decisions, especially around security, are still being made using no or weak data,” he says. “So much is based on guesswork and gut feeling. It has to get better.”
Key to A.I.’s impact on business is the ability for the data to be collected, analyzed and acted upon quickly. “We are moving from a reactionary world to a proactive one,” says Zahran. “Using A.I. to harness vast data sets and quickly analyze multiple variables provides an empirical perspective from which intelligent decisions can be made. A.I. provides a look at causality that just isn’t available when relying solely on human analysis.”
A.I. puts data into context and analyzes the impact. For example, a customer gets an alert about potentially fraudulent activity based on a purchase at XYZ Store. Meanwhile, the credit card company’s system analyzes recent transactions at that particular location and determines that several fraud cases have popped up in the past few weeks, triggering an alert to the store’s loss prevention team. The team then reviews the data and determines if and finally which employees are perpetrating the fraud, preventing future losses. This type of quick analysis and action could not be possible without A.I.
“It allows companies to cast a much wider net while also being able to boil data down to a very micro level perspective quickly,” say Zahran. “It really is game changing for our industry.”
Changing the profile of risk management
Artificial intelligence. Machine learning. Big data. Pinkerton’s Applied Risk Science. It all requires a much different level of understanding about how technology and data flow are impacting global companies.
“A.I. has broken down the traditional boundaries as everything within an enterprise, and beyond, becomes more connected,” says Zahran. “It is leading to more intelligence that can be used to mitigate so many risks companies face today. With silos gone, security personnel have to adapt and understand much more about their enterprise.”
As a result, the profile of a risk management professional is evolving. The industry, once mostly staffed by former law enforcement, government and military professionals, is now looking at data specialists and analysts as key personnel. “It’s another way that A.I. and machine learning have impacted our industry,” says Williams. “No longer are companies content with basic protections. They require the ability to prevent incidents that could have huge potential impacts on their business. Data is at the center of that, which requires understanding that language.”
People are still the key, however. “Enterprise-wide decisions are still made by people,” explains Zahran. “Machines learn based on the data and present cause-based analysis but in the end, it is people who decide the next courses of action. As we continue to utilize the latest technology in our business, it’s critical that our people understand how to take an empirical approach to the proactive recommendations they make for our clients.”
