What counts as a data leak
A data leak (data breach) is a situation in which the organization’s information reaches someone who is not supposed to see it. The information can be personal, such as customer details, ID numbers, medical information or payment details, or business information: price quotes, contracts, code or financial data. Under the Privacy Protection Law, the central question is whether personal data held in one of the organization’s databases was exposed.
A leak does not have to start with an outside attacker. It can come from an attack, such as a server break-in or an account takeover, from an insider who copies information before leaving, or from a simple mistake: a file sent to the wrong recipient, a cloud folder shared with an open link, or a database mistakenly configured as reachable from the internet (misconfiguration). In every case the result is similar: the information is out, and it cannot be brought back.
What raises the risk
- Overly broad permissions: employees who can access every folder and every database, even when their role does not require it, contrary to the principle of least privilege.
- Data whose location is unknown: there is no mapping of the databases (data mapping), so there is no way to know what needs protecting and what exactly was exposed.
- Uncontrolled sharing: open sharing links, files forwarded to private mailboxes, unapproved cloud services (Shadow IT), and employees using personal AI tools (Shadow AI) to which customer documents are uploaded.
- Data kept too long: old databases, exports and copies nobody needs anymore but that can still be stolen. A data retention policy sets what is deleted and when.
- Cloud configuration mistakes: storage, databases or test environments opened to public access.
- Unencrypted data (encryption): laptops, drives and backups that anyone holding them can read.
- Employees leaving without a procedure (offboarding): accounts that stay open, and access to data that is never closed.
Early warning signs
A leak is often discovered from outside rather than from within the organization’s systems. These are signs to take seriously:
- Customers who mention fraudulent messages that know details only the organization had.
- Contact from a security researcher, a journalist or an official body, including the National Cyber Directorate, about the organization’s information found online.
- The organization’s documents appearing somewhere unexpected.
- Unusual downloading or copying of files, especially before an employee leaves or at unusual hours.
- Alerts from protection systems about data being sent out (data exfiltration), or about external sharing of sensitive folders.
- A threat or a demand for payment from someone who claims to hold the organization’s data.
The impact on the organization
Data that has left does not come back. Even if the breach was closed quickly, the customers whose data was exposed may be targets of fraud and impersonation long after the incident. They will contact the organization with questions, some will demand compensation, and others will move to another provider. In a business built on trust, such as an accounting firm, a clinic or an insurance agency, this damage can be greater than any technical cost.
Alongside the loss of trust there is legal exposure: customer lawsuits, claims by business partners whose data was affected, and an inquiry by the Privacy Protection Authority. When the leak comes with an extortion demand, time pressure is added too, and under that kind of pressure it is easy to make wrong decisions.
The regulatory side: the Privacy Protection Law and Amendment 13
This section is general information and is not legal advice. In any incident of this kind it is worth involving a lawyer who specializes in privacy.
The Privacy Protection Law and the Privacy Protection Regulations (Data Security), 2017 require an organization that holds a database of personal data to secure it at a level that fits the sensitivity of the data. The regulations also set a duty to report a serious security incident to the Privacy Protection Authority (breach notification), depending on the database’s security level and the type of incident. In some cases the Authority may instruct the organization to also notify the people whose data was exposed.
Amendment 13 to the law, which came into force in August 2025, expanded the Authority’s supervision and enforcement powers, including the imposition of financial sanctions. In practical terms, after a leak the organization needs to be able to show that it took reasonable security measures, that it has a mapping of its databases, and that it handled the incident according to a procedure. So as soon as the incident is discovered, check whether it meets the definition of a serious incident and whether there is a duty to report. This decision is made quickly, based on facts, and documented.
How to reduce the risk
- Data mapping and classification (data classification): knowing which databases exist, what they contain and who accesses them. This is also the basis for meeting the regulations.
- Periodic permission review (access review): who accesses which data, and whether they still need that access.
- Encryption of laptops, backups and sensitive data in the cloud.
- Data leak prevention controls (DLP): an alert or a block when sensitive data is sent out or shared through an open link.
- A policy for sharing and tool use: which services are approved, what may be shared with outside parties, and what must not be uploaded to AI tools.
- Data minimization: not collecting what is not needed, and deleting what is no longer required.
- Security monitoring and alerting (SIEM / SOC): tracking unusual access to data and large-volume copying.
- A ready incident procedure (incident response plan): who decides, who checks the reporting duty, and how to talk to customers.
When data has leaked: the first steps with the IR team
As soon as a leak is suspected, it is worth bringing in a cyber incident response team, and at the same time the organization’s management and the legal advisor. These are the main steps:
- Stopping the leak (containment): closing the sharing link, locking the compromised account, blocking public access to the database or disconnecting the server.
- Preserving evidence (forensics): logs, messages and copies of what was exposed, before anything is changed or deleted.
- Assessing the scope: what data got out, whose, how sensitive and over what period. These answers are the basis for the reporting decision.
- Reviewing the reporting duty: to the Privacy Protection Authority, and as needed also to the insurance company, business partners or the National Cyber Directorate’s 119 hotline.
- Communicating with customers: a factual, accurate message that explains what happened and what they should do, for example watch for fraudulent messages that use their details.
- Lessons learned: fixing the cause, whether it is a wrong configuration, broad permissions or a missing offboarding procedure.
When the leak comes with a ransom demand, there is no certainty that paying will prevent publication of the data, and no way to verify that the data was really deleted. The decision is made together with the IR team, the legal advisor and the insurance company, not under the pressure of the first few minutes.
Summary
- A data leak can come from an attack, an insider or a configuration mistake, and the result is similar in every case.
- Broad permissions, uncontrolled sharing and unmapped data are the factors that raise the risk.
- Under the Privacy Protection Law and Amendment 13, the duty to report to the Privacy Protection Authority must be reviewed immediately, and the decision documented.
- When data has leaked, stop the leak, preserve evidence, assess the scope and bring in the IR team and a legal advisor.