Insider Threats: Are Disgruntled Employees a Cybersecurity Risk?

Limiting access, creating protocols aimed at keeping data safe can mitigate sabotage.

Most plan sponsors’ cybersecurity concerns are that outside hackers will attempt to get access to their systems, but disgruntled employees can also pose a threat.

The Department of Labor recently updated its cybersecurity guidance to cover all Employee Retirement Income Security Act employee benefit plans, including health and welfare plans, along with retirement plans. That means plan sponsors have much more data to protect.

Never miss a story — sign up for PLANSPONSOR newsletters to keep up on the latest retirement plan benefits news.

Fortunately for plan sponsors, employees do not have broad access to critical information such as individual participant account passwords or retirement funds, which sit with recordkeepers and custodial banks, respectively, says Sean Fullerton, a senior investment strategist for the defined contribution team at Allspring Global Investments.

Internal threats account for about 20% of security threats, according to the Verizon 2022 Data Breach Investigations Report, making them rarer than outsider cybersecurity hacks. Jenny Eller, a principal in Groom Law’s retirement services practice group, says in the 25 years she has been practicing law, she had a single anecdote of an employee who tried to commit fraud by creating a dummy account.

Still, certain employees, such as those in human resources, information technology or treasury, may have access to plan information or other personally identifiable information. There are, however, ways to prevent or limit potential damage caused by disgruntled employees.

Limiting Access

The DOL lists a dozen best practices plan sponsors should utilize to protect their employees; chief among them is limiting access to the plan administration. It is also important plan sponsors have written and documented internal control policies, says Julie Doran Stewart, head of fiduciary advisory services at Sentinel Group, an adviser and recordkeeper to plans. Those policies include the steps a company’s HR or IT team needs to take to shut off the access for someone at the organizational level, but they also include timely communication with vendors.

“If we have a client that doesn’t tell us that this happens, then we’re only as good as the information we have,” she says.

Sentinel occasionally audits who is listed as having access to the plans they advise.

“It’s being diligent about double-checking the access points and doing sort of an internal audit, if you will, on both sides, on a periodic basis,” Doran Stewart says.

Fullerton says plan sponsors should also talk to their service providers to understand what standards of information security the providers use and validate the providers’ processes to ensure the plan sponsor is comfortable with how the organizations handle cybersecurity.  

Doran Stewart says checking with vendors should be done periodically and can be as simple as sending an annual due diligence questionnaire to advisers, recordkeepers and third-party administrators to confirm their cybersecurity policies, including specifically asking about access controls.

“The Department of Labor obviously has made this a priority from a fiduciary governance perspective, so they are going to be looking for procedures and records related to that due diligence being done,” Doran Stewart says.

The more people who have access to data or accounts, the more risk there is for fraud, says Tim Rouse, the executive director of the SPARK [Society of Professional Asset Managers and Recordkeepers] Institute, which created many of the cybersecurity best practices the DOL shares.

At the adviser level, Rouse says SPARK is leery of individuals giving advisers access to their accounts and allowing tools like screen-scraping capabilities, which have the potential for abuse.

“Other than communicating those concerns, those decisions come either at the plan sponsor level or at the individual participant’s level,” he says.

Using Technology

Plan sponsors should encourage employees to regularly log into their accounts to keep them secure, especially if it has been more than a year since they have logged in, and to use security tools such as multi-factor authentication, Doran Stewart says. That is especially important if the plan changed recordkeepers, as inactive accounts can be hacked. There may be a misconception among plan participants that if they never log in, the account will be safer, but not setting up controls makes it easier for a bad actor to hack, she adds.

There are opportunities for disgruntled employees at the plan sponsor level to embezzle money before it gets into the account, and internal controls such as audits can also help keep employee funds secure, Rouse says.

Rouse adds that SPARK is working with a committee of third-party administrators to create common file formats for Application Programming Interface, or API, connectivity to establish a more streamlined and efficient way to send data, compared with emailing spreadsheets. An API is also more resilient against cyberattacks.

Using detective controls—which can help spot and respond to security issues—on data usage can alert information technology departments to any unusual data activity, such as someone logging in at odd hours or making large downloads, says Lou Steinberg, founder of and managing partner in CTM Insights LLC, a cybersecurity research lab and funder.

Many employee benefit plans can be accessed using different methods, such as phone, computer or mobile apps, so plan sponsors should make sure both their benefits team and IT department participate in the due diligence process when they meet with vendors.

“Those are two different skill sets … so keeping an open line of communication [about] how they can mutually assist each other … is important,” Doran Stewart says.

A Guide to Buying and Maintaining Cyberinsurance

Now more than ever, plan sponsors need to understand what types of coverage are available and what will fit them best.

This story has been updated for the magazine. That version can be found here: “A Guide to Buying and Maintaining Cyberinsurance

Employee benefits and retirement plans are a natural target for cybercriminals. Allie Itami, a partner in the law firm Lathrop GPM LLP, notes that health and welfare and retirement plans have large amounts of valuable data, including personal or sensitive information about employees and their beneficiaries.

Never miss a story — sign up for PLANSPONSOR newsletters to keep up on the latest retirement plan benefits news.

Plan regulators recognize the problem, as evidenced by the Department of Labor’s recent cybersecurity guidance update, compliance assistance release 2024-01, which includes, “You may want to require insurance coverage such as professional liability and errors and omissions liability insurance, cyber liability and privacy breach insurance, and/or fidelity bond/blanket crime coverage.”

Does Your Plan Need Cyberinsurance?

Itami says the DOL guidance assumes that insurance is available for cybersecurity breaches or incidents but does not specify cyberinsurance as the only insurance type. Cyberinsurance covers loss and costs associated with a data breach but does not typically offer coverage related to the theft of money.

“The loss of account balances is more likely to be covered by a crime- or fiduciary liability policy,” Itami notes. “So simply purchasing a cybersecurity policy for a plan may not provide all the coverage desired.”

Carol Buckmann, a partner in Cohen & Buckmann P.C., says plan sponsors should understand that their fiduciary liability policy is not a substitute for cyberinsurance. State and other laws impose cybersecurity obligations, so there may not be any allegations of Employee Retirement Income Security Act fiduciary breach made when there is a security breach. As the DOL’s release noted, welfare plans’ fiduciaries have ERISA obligations to keep plan assets and data safe, even though they may also be subject to HIPAA’s Security Rule. “There was confusion about that, since prior DOL guidance implied it but didn’t clearly spell it out,” says Buckmann.

A general cyberinsurance policy’s coverage can be limited. Itami says that although there could be a carve-out for ERISA plans in a general cyberinsurance policy, there is more likely an ERISA fiduciary breach exclusion. She says this situation creates a question of whether a cybersecurity breach is an ERISA fiduciary duty breach, triggering the exclusion.

Buying a Policy

The cyberinsurance market is intricate, according to Richard Clarke, chief insurance officer of Colonial Surety Co. Some product sellers are more knowledgeable than others, and some have access to more potential markets. Cyberinsurance is not standardized, so sponsors need basic knowledge to evaluate insurance options and policy details properly.

Buckmann explains that cyberinsurance is not like an ERISA bond, which is relatively standard across insurers. She says different types of digital coverage are available: “For example, some policies may cover only employees of the plan sponsor, while others may cover third parties.”

A November 2023 report from tax and advisory firm BDO USA recommended that plan sponsors considering retirement plan cybersecurity insurance ask which party—the plan or the sponsor—would be liable for a cybersecurity breach. The report suggested other points to review, including “identifying who is the insured party (the sponsor, the plan or both?), who is responsible for purchasing the policy (the sponsor or the plan?), and the full scope of the policy (in other words, what is or is not covered in the event of a cyber breach?).”

“It is preferrable to remove doubt by naming a plan as a covered entity and seeking explicit coverage for ERISA fiduciary breaches, as well as nonfiduciary cybersecurity breaches,” Itami cautions. “Additionally, if the ERISA plan is the only insured, the fiduciary might more easily conclude that use of plan assets for obtaining the policy is appropriate.”

Clarke says insurance underwriters are becoming more rigid in their decisionmaking, relying heavily on applications and supplemental information before issuing coverage.

“For example, it is almost mandatory for insurance applicants to have multi-factor authentication in place as a prerequisite for obtaining cyberinsurance,” he says. “Some underwriters even require more extensive internal protections than just MFA, depending on the applicant’s risk characteristics.”

In Buckmann’s experience, plan sponsors often wonder how much cybersecurity coverage to buy. She cites one expert whose stock answer is to “buy as much as you can afford.” The reasoning is that a cybersecurity breach has many potentially expensive consequences. “These can include costs of a breach response, ransomware payments, business interruptions or reputational harm, losses from cybercrime and liabilities to third parties,” says Buckmann. “Plan sponsors should not underestimate their potential exposure in deciding how much insurance to purchase.”

Maintaining a Policy

A 2023 study sponsored by Recast Software and conducted by Ponemon Institute, a Michigan-based research center dedicated to privacy, data protection and information security, supports Clarke’s impression of more rigorous underwriting: 50% of the participating information technology and security respondents said it was difficult or very difficult to comply with their insurer’s requirements. More than half reported that their insurer required regular scanning for system vulnerabilities; 43% reported a requirement to scan multiple times each week.

Insurers will want to see plans implement steps recommended by the DOL and other sources of best practices, says Buckmann. Those sources could include the best practices framework of the National Institute of Standards and Technology or the SPARK [Society of Professional Asset Managers and Recordkeepers] Institute.

“The types of practices they will want to see include annual third-party audits of their systems, encryption of sensitive data, employee training, monitoring security of remote workers, good controls on access to data and service provider reviews and assessments,” Buckmann adds. 

«