Trends in Risk Management

Hacking the Human Brain: How Cognitive Science Helps Mitigate Risk

New technologies drawing on behavioral science know when you're about to make a mistake.
By: | July 18, 2018 • 4 min read

Human behavioral tendencies and individual decision-making factor largely into any organization’s risk profile.


The urge to look at a flashy billboard distracts a truck driver just long enough to cause an accident. Simple forgetfulness causes a retail worker to neglect to set his store’s alarm system when he closes up for the night. The desire to act quickly on an urgent request from the boss drives an office worker to wire a sum of money without verifying the request.

All of these losses are avoidable, but even the most comprehensive employee training can’t re-wire the brain to prevent mistakes 100 percent of the time.

Enter Behavioral Science

According to a recent Deloitte report, “The future of risk,” behavioral science is “the study of human behavior through systematic research and scientific methods, drawing from psychology, neuroscience, cognitive science and the social sciences.” Businesses are turning to its methods to identify the thought processes and biases that drive people to make risky decisions, and how those behaviors could be modified.

“There are two levels where behavioral and cognitive science can be applied to risk management. The first is making normal risk management processes more efficient and effective. The second is, more broadly, allowing decision-makers to better identify opportunities worth pursuing,” said Dilip Krishna, chief technology officer and a managing director with the Regulatory & Operational Risk practice at Deloitte & Touche LLP.

Krishna said natural language processing technologies can be used to apply the principles of cognitive science to employee behaviors. According to analytics software provider SAS, “natural language processing is a branch of artificial intelligence that helps computers understand, interpret and manipulate human language.”

A scan of employee emails might, for example, be able to identify when a worker is overly stressed just by the tone or language he uses.

This includes data that exists in Word documents, Excel spreadsheets, emails, texts and images — formats workers use to communicate but which computers do not easily interpret. Natural language processing can recast that data so computers can better understand it, allowing risk managers to tap into broader insights.

A scan of employee emails might, for example, be able to identify when a worker is overly stressed just by the tone or language they use.

“People under stress are more prone to making mistakes,” Krishna said. That insight offers an opportunity to intervene and head off mistakes proactively.

Predicting Employee Cyber Risk

Nowhere might this technology be more relevant than in the realm of cyber risk — specifically in the form of social engineering.

This form of theft or infiltration is built around manipulation of the mind and so offers the greatest potential for cognitive science to make improvements.

Tokyo-based technology company Fujitsu has already put this in play.

An enterprise-wide platform uses psychological profiling to identify the employees most likely to be tricked by a fraudulent email or link, based on how often they click links from emails, how quickly they respond to emails and their behavior while browsing the web. The organization can then provide additional, targeted training for those employees.

Pros and Cons of Cognitive Science Tech

But some employees might perceive those interventions as an unwarranted or unfair overstep. According to the Deloitte report, employees might “see behavioral interventions as an impingement of free will,” and organizations face the “risk of regulatory action in case of perceived misuse of behavioral intervention.”

“The world was built by humans for humans, but if we can take the way that computers think and apply that type of thought processing to the way we work, we can unlock a lot of insights.” — Dilip Krishna, chief technology officer and a managing director, Deloitte & Touche LLP

Still, Krishna said the long-term benefits of such technology from a risk management standpoint could be significant.

“The world was built by humans for humans, but if we can take the way that computers think and apply that type of thought processing to the way we work, we can unlock a lot of insights,” he said. “There’s potential for anywhere from 20 to 90 percent improvement from a risk profile perspective.”


In addition to employee and regulatory pushback, slow return on investment is also a roadblock. Despite its potential to reduce risks, Krishna acknowledged that parsing through complex human thought processes devising successful interventions takes time to get right.

But behavioral science has less-controversial applications in risk management as well.

“Design thinking” can train executives to overcome personal bias in decision-making, in pursuit of a more objective and thoughtful view of new business opportunities. This type of behavioral intervention helps business leaders take advantage of the upside of risk rather than avoid the downside, addressing the “second level” of risk management, as Krishna described.

As the amount of data grows exponentially and AI becomes even smarter, applications of cognitive science certainly have a place on risk managers’ radar. &

Katie Dwyer is an associate editor at Risk & Insurance®. She can be reached at [email protected]

More from Risk & Insurance

More from Risk & Insurance

4 Companies That Rocked It by Treating Injured Workers as Equals; Not Adversaries

The 2018 Teddy Award winners built their programs around people, not claims, and offer proof that a worker-centric approach is a smarter way to operate.
By: | October 30, 2018 • 3 min read

Across the workers’ compensation industry, the concept of a worker advocacy model has been around for a while, but has only seen notable adoption in recent years.

Even among those not adopting a formal advocacy approach, mindsets are shifting. Formerly claims-centric programs are becoming worker-centric and it’s a win all around: better outcomes; greater productivity; safer, healthier employees and a stronger bottom line.


That’s what you’ll see in this month’s issue of Risk & Insurance® when you read the profiles of the four recipients of the 2018 Theodore Roosevelt Workers’ Compensation and Disability Management Award, sponsored by PMA Companies. These four programs put workers front and center in everything they do.

“We were focused on building up a program with an eye on our partner experience. Cost was at the bottom of the list. Doing a better job by our partners was at the top,” said Steve Legg, director of risk management for Starbucks.

Starbucks put claims reporting in the hands of its partners, an exemplary act of trust. The coffee company also put itself in workers’ shoes to identify and remove points of friction.

That led to a call center run by Starbucks’ TPA and a dedicated telephonic case management team so that partners can speak to a live person without the frustration of ‘phone tag’ and unanswered questions.

“We were focused on building up a program with an eye on our partner experience. Cost was at the bottom of the list. Doing a better job by our partners was at the top.” — Steve Legg, director of risk management, Starbucks

Starbucks also implemented direct deposit for lost-time pay, eliminating stressful wait times for injured partners, and allowing them to focus on healing.

For Starbucks, as for all of the 2018 Teddy Award winners, the approach is netting measurable results. With higher partner satisfaction, it has seen a 50 percent decrease in litigation.

Teddy winner Main Line Health (MLH) adopted worker advocacy in a way that goes far beyond claims.

Employees who identify and report safety hazards can take credit for their actions by sending out a formal “Employee Safety Message” to nearly 11,000 mailboxes across the organization.

“The recognition is pretty cool,” said Steve Besack, system director, claims management and workers’ compensation for the health system.

MLH also takes a non-adversarial approach to workers with repeat injuries, seeing them as a resource for identifying areas of improvement.

“When you look at ‘repeat offenders’ in an unconventional way, they’re a great asset to the program, not a liability,” said Mike Miller, manager, workers’ compensation and employee safety for MLH.

Teddy winner Monmouth County, N.J. utilizes high-tech motion capture technology to reduce the chance of placing new hires in jobs that are likely to hurt them.

Monmouth County also adopted numerous wellness initiatives that help workers manage their weight and improve their wellbeing overall.

“You should see the looks on their faces when their cholesterol is down, they’ve lost weight and their blood sugar is better. We’ve had people lose 30 and 40 pounds,” said William McGuane, the county’s manager of benefits and workers’ compensation.


Do these sound like minor program elements? The math says otherwise: Claims severity has plunged from $5.5 million in 2009 to $1.3 million in 2017.

At the University of Pennsylvania, putting workers first means getting out from behind the desk and finding out what each one of them is tasked with, day in, day out — and looking for ways to make each of those tasks safer.

Regular observations across the sprawling campus have resulted in a phenomenal number of process and equipment changes that seem simple on their own, but in combination have created a substantially safer, healthier campus and improved employee morale.

UPenn’s workers’ comp costs, in the seven-digit figures in 2009, have been virtually cut in half.

Risk & Insurance® is proud to honor the work of these four organizations. We hope their stories inspire other organizations to be true partners with the employees they depend on. &

Michelle Kerr is associate editor of Risk & Insurance. She can be reached at [email protected]