A survey on Data Loss Prevention, being conducted in partnership with Nclose and Tessian, has gone live on ITWeb. The objective of the survey is to find out the measures that South African organisations are taking to prevent loss of data and the role of machine learning in data loss prevention (DLP).

Stephen Osler, co-founder and business development director at Nclose, says: “With the recent implementation of the POPIA legislation, the survey aims to provide some visibility and clarity into potential control mechanisms that could be implemented to help prevent loss of personal information.

“The reality is that the majority of data privacy providers are trying to leverage POPIA as a motivation for businesses to buy their solutions. However, in the more than a decade that we’ve been involved in DLP projects, it’s become apparent that traditional DLP solutions generate so many false positives, often resulting in alert fatigue, that businesses eventually turn them off,” says Osler.

To this end, the survey also aims to establish whether respondents agree that what’s needed is a data privacy product that doesn’t add unnecessary noise in their environment.

“Another concern is that businesses are approaching POPIA compliance as a box-ticking exercise instead of deploying a solution that’s fit for purpose,” says George Vasey, partnerships manager at Tessian.

“Stereotypical data leakage prevention products look at all of the requisite data protection policies and tick all of the relevant boxes, but what the business ends up with is static policies that look at static rules of engagement," adds Olser, meaning that the majority of DLP solutions generally don’t understand the context in which the person is engaging with the data.

“There’s no historical context that looks at how the person engaged with the data previously, so the solution isn’t able to understand changes in how someone engages with data,” says Vasey.

“With this survey we want to understand the respondents’ appetite for a DLP solution that uses machine learning to identify variations in how someone is engaging with data, whether it be accidental or malicious," Vasey explains. "The aim is to avoid triggers for normal business use, just alerting for malicious or accidental variations and to use those accidental variations as learning tools to educate the end user on how to engage with the businesses data securely. We’ve been talking machine learning when it comes to everywhere else in the security space except for people’s interactions with data.”

Osler further adds, “Over the past five years a lot of vendors have mentioned using analytics to assess user behaviour, but we haven’t yet seen an actual solution that works without requiring cumbersome products that don’t justify the end result.”

The survey asks respondents whether the POPIA legislation has made them reconsider their DLP. It also investigates if they think that leveraging machine learning and behavioural intelligence could provide better data loss prevention, and whether they’re most concerned about accidental data loss or intentional data exfiltration.

We hope you’ll be able set aside a few minutes of your time to participate in the survey, and stand a chance to win a lucky draw prize, a Takealot voucher to the value of R3 000

The detailed results of the survey, and the prize winner, will be published on ITWeb.

To play your role in compiling this data loss prevention trends report, follow this link.

What are the best control measures to prevent data loss?