By Guy Eisdorfer

When a company announces layoffs, one of the last things most employees or even company owners worry about is data loss.

Related: The importance of preserving trust in 2023

Valuable or sensitive information on a computer is exposed to theft or to getting compromised. This can happen due to intentional theft, human error, malware, or even physical destruction of servers. But it’s a real and growing risk to be aware of.

In 2020, Forbes reported that pandemic layoffs and remote work served to increase the risk of company data loss. Tesla, for example, suffered two cybersecurity events after layoffs back in 2018.

Data loss isn’t necessarily spiteful. Imagine an employee creates a spreadsheet showing all your clients and the main points of contact for each. She updates this sheet, but forgets to share it internally.

She gets laid off, and she takes the spreadsheet with her because she believes that the work she created at her job belongs to her. This may sound like an edge case, but a survey by Biscom found that 87 percent of employees took data that they themselves had created from their last job.

Data theft can also be deliberate and malicious. That same employee might use that spreadsheet as a bargaining chip in securing a new job with your competitor.

Data theft can also happen as a result of hackers. In the infamous 2014 Sony hack, an employee moving from Deloitte to Sony allegedly took sensitive data with him when he left. It is believed that the employee was storing employee information from both Sony and Deloitte in his computer, leading to the salaries of 30,000 Deloitte employees being leaked.

Data loss prevention is a concept that’s been around since the ‘90s, but in the age of AI, machine learning, natural language processing, and all those other fun new buzzwords, it’s taken on new relevance and significance.

With relaxed security measures due to remote work, disgruntled employees due to sudden mass layoffs, and logistical oversights due to reorganization, company data can fall through the cracks. To keep up, companies need to use technology to ensure their most important asset, their information, is safe.

Consolidated visibility

The first step is to know what you have. Then you can work on protecting it.

That’s why the first step in any layoff-proof data loss prevention strategy has to be the collection and categorization of all the company data that exists. This is both easier and harder thanks to a distributed system of information.

Data might be in spreadsheets, on Slack, on OneDrive, in custom databases, or any other number of off-premises cloud systems.

The best way to consolidate all that info is to use machine learning and artificial intelligence. First, identify all potential sources of data. You might also want to ensure you’re scanning all emails going in and out of the company.

Then, companies need to set up rules to determine what the AI identifies as what kind of data. For example, one priority is identifying personally identifiable information of your customers. You don’t want that leaving your data warehouses.

Another example is any kind of proprietary algorithm or system. For instance, if you’re Equifax, you don’t want any employee able to leave with your credit score algorithm.

Using a combination of AI and ML, you should be able to put together a comprehensive catalog of all company data.

Spotting anamolies

The next step is to train the AI to spot suspicious-looking behavior. For example, you might set it up so that when an employee starts downloading massive amounts of data, that gets flagged as suspicious.

You might also need to use technology that can use optical character recognition (OCR). For example, imagine instead of sharing that customer spreadsheet, our laid-off employee just takes a screenshot of it and emails it to herself.

Unless your data loss prevention strategy has OCR to read what screenshots are, you’d never be able to know that she walked off with that spreadsheet unless you manually went through every single one of her emails.

You also have to take steps to stop data loss from happening. For example, your system should include a rule to automatically log out any users downloading a high number of files. It should also limit access for any soon-to-be laid off employees to sensitive material.

And finally, in the case of non-malicious theft, you should be able to quickly scan any employee-generated data to ensure files like comprehensive customer databases don’t get lost just because nobody knows they exist.

One major component of data loss prevention is to map the organization’s critical information. With a map of who has access to what, the knowledge is less likely to get lost when employees move on. This enables companies to classify the information and prevent data loss, or at least educate employees not to take data with them to their next job.

You should also have set up your system to flag suspicious events, such as the mass downloading of files, laid-off employees sending lots of emails, or people logging in from unusual locations.

Your final step is to patch those holes. With AI on the case, it will auto-recognize suspicious events and take care of them. You can also be assured that important or sensitive information won’t fall through the cracks of mass layoffs.

Data loss is a real threat. Make sure your company is up to the job of handling it.

About the essayist:Guy Eisdorfer, is the co-founder and CEO of Cognni, a supplier of AI-powered data classification systems  and other security products to enterprises and SMBs.

February 13th, 2023