How can machine learning prevent data loss from emails?

More companies are experiencing data loss from emails, but how can machine learning help?

How can machine learning prevent data loss from emails?

According to Tessian, data loss from emails remains the biggest threat facing digital businesses today. In order to diminish the risks and problems associated with email, the company offers a number of innovative solutions.

Stopping misdirected emails

In effect, misdirected emails are a product of a sender accidentally inputting the incorrect recipient. As a result, this has the potential to mistakenly disclose highly sensitive information to the wrong person.

As Tessian highlights, the volume and sensitivity of data shared via email can cause “significant damage” to a company. This includes the loss of customers and revenue, alongside fines and penalties issued by regulators and long-term reputational damage.

According to Tessian, a 10,000 person organisation has the potential to suffer from 7,000 misdirected emails per year. In fact, the Information Commissioner’s Office reported misdirected emails as the most common data security incident in 2017.

Inadvertent human error is an unfortunate reality for most companies. Nevertheless, Tessian Guardian‘s “stateful machine learning” automatically prevents wrong recipients from receiving “highly sensitive emails.”

In order to do, the tool uses natural language processing and the Tessian Parallax Engine. This automatically “analyses all outbound emails in real-time before warning users if something looks unusual.”

Preventing unauthorised emails

Another integral element of email protection is preventing unauthorised emails. This entails an employee sending highly sensitive emails to their personal email account or unauthorised email accounts for “well intentioned or malicious reasons.”

If employee’s share sensitive information with an unauthorised recipient through email, this discloses data to a third party. As Tessian observes, this valuable data now resides on email servers out of the enterprise’s control.

These events can therefore breach customer contracts, such as non-disclosure agreements or data protection regulations. For a 10,000 person company, the number of potential breaches per year due to unauthorised emails is 195,000.

In addition, 69% of organisations polled reported significant data loss due to employees taking information with them when leaving the business. In order to prevent unauthorised emails, Tessian Enforcer stops sensitive emails being sent to unauthorised and personal email accounts.

The Tessian Parallax Engine and natural language processing automatically analyses all outbound emails in real-time. Companies can therefore benefit from automatic protection, reduced data loss from emails, and peace of mind.

How can companies ensure data protection and recovery? Owner of FreedomFire Communications, Michael Lemm, offered his insights