710+ million
records of Russian citizens leaked in 2024
Roskomnadzor data: 135 incidents during the year. InfoWatch estimates 1.58 billion records, up 30% from 2023.
Built in-house
DataSafe — Personal Data Protection
Protection and anonymisation of personal data in CRM, HR and other corporate systems. DataSafe hides sensitive fields, controls disclosure and logs access without replacing business systems.
#security#compliance#data-protection#ai-monitoring
- Scope
- CRM, HR and related systems
- Deployment
- Local, without a public perimeter
- Regulatory framework
- Federal Law No. 152-FZ, Government Decree No. 1154, Federal Law No. 420-FZ
- Licensing
- Based on the number of environments and controlled users
/ Interface and video
13 materials/ The cost of doing nothing
Personal-data protection after stronger liability from 30 May 2025
Federal Law No. 420-FZ introduced turnover-based fines and a scale based on the size of a leak. At the same time, Article 272.1 of the Criminal Code applies, with up to ten years of imprisonment.
A data leak is no longer an IT incident that can be resolved quietly. The legal, financial and reputational consequences now reach the board level. DataSafe helps reduce risk before the data leaves the company: sensitive values are separated, access is controlled, and every disclosure is logged.
Violation scope
Before May 30, 2025
From May 30, 2025
Leak of 1,000–10,000 data subjects / 10,000–100,000 identifiers
₽60,000–100,000
₽3–5 million
Leak of 10,000–100,000 data subjects / 100,000–1 million identifiers
₽60,000–100,000
₽5–10 million
Leak of over 100,000 data subjects / over 1 million identifiers
₽60,000–100,000
₽10–15 million
Leak of special-category data (medical data, political views)
₽60,000–100,000
₽10–15 million
Leak of biometric data
₽60,000–100,000
₽15–20 million
Failure to notify Roskomnadzor of a leak within 24 hours
—
₽1–3 million
Repeated leak in any category
₽60,000
1–3% of annual revenue, up to ₽500 million
— Main point —
Protection works when it is built into everyday work, not added after an incident. DataSafe does not replace business systems: it adds a controlled layer around personal data so people keep working in familiar tools while the company gains visibility and evidence.
66%
of incidents involve employee actions
Staff errors and misuse of legitimate access, rather than external attacks. Perimeter tools do not detect such cases.
₽575 million
damage from three days of downtime at CDEK
Head Mare attack, May 2024. Systems were encrypted and backups destroyed. Experts estimate total losses at ₽300 million to ₽1 billion.
up to 10 years
of imprisonment under Article 272.1
Effective from 11 December 2024. Illegal transfer, collection and storage of personal data with serious consequences.
/ Three leak paths
Where personal data really leaks
DataSafe closes three common leak scenarios that perimeter protection and a traditional SIEM cannot cover on their own.
01External attack
A hacker enters the infrastructure and exports the database
A group enters through a vulnerability, phishing or a compromised contractor account. Then it exports tables with real personal data or encrypts systems and demands payment. Perimeter security detects the attack, but by then the data may already be gone.
— Practice example —
Real cases: Gemotest, April 2022; CDEK, May 2024
At Gemotest, an employee account was compromised and 300 GB of data from up to 30 million customers was exported, including medical test results. At CDEK, a ransomware attack disrupted systems and destroyed backups. The issue is not only entry into the infrastructure but the value of the data available after entry.
How DataSafe — Personal Data Protection closes it
- 01Sensitive fields are stored separately from the production database and are available only through DataSafe for targeted requests.Even after a full table export
- 02the attacker receives anonymised values rather than real personal data.Access is controlled by role
- 03scenario and signed requests
- 04and every disclosure is logged.
02Insider
An employee or contractor with legitimate access
The user has access rights required for their role. They export data manually, through bulk reports or through an API. A traditional SIEM sees a legitimate login and does not know whether the request was justified.
— Practice example —
Real cases: Sberbank, MTS Bank, Ozon Bank and Rosselkhozbank
A Sberbank department head exported and published a customer database. An MTS Bank branch manager sold data on 5,600 customers to fraudsters and received three and a half years in prison. In these cases, the attacker already had legitimate access.
How DataSafe — Personal Data Protection closes it
- 01By default
- 02users see protected valuesrather than real data. Disclosure is allowed only in an explicit
- 03signed scenario.Every access to a sensitive field is logged with role
- 04purpose and time.Behavioural analytics identifies unusual volumes
- 05times and patterns and sends alerts to existing security tools.
03Test environments
Real data in test, demo and development environments
Real personal data is copied from production to a test environment so the team can verify functionality. Contractors, analysts and demo environments gain access. These environments usually have lower protection, which makes them a convenient leak channel.
— Practice example —
A common scenario and Government Decree No. 1154 of 30 May 2025
This is one of the most common leak channels, created by companies themselves. Government Decree No. 1154 directly requires personal-data anonymisation when information systems are used for testing, training and demonstrations.
How DataSafe — Personal Data Protection closes it
- 01Testtraining and demo environments receive anonymised data sets instead of real values. Structure and formats are preserved.Contractors and analysts work with the data model and do not gain access to personal data.DataSafe controls export
- 02creation and refresh of each data set.
/ What the solution does
Personal-data protection and anonymisation in corporate systems
Four functions that create a protected personal-data environment inside your systems—without replacing CRM or rewriting product code.
01
Hides sensitive fields completely
By default, users see protected fields rather than the values themselves. Disclosure depends on role and scenario and requires an explicit employee action.
02
Anonymises test and development environments
Test, demo and training environments receive anonymised sets instead of real personal data. Structure, types and relationships are preserved, so functionality is unaffected.
03
Logs requests without recording personal data
Every data request becomes traceable, but personal data itself never enters the log. The log provides context, not a second copy of the database.
04
Detects deviations and notifies security
Suspicious scenarios are identified through behavioural analytics and passed to existing monitoring and response tools: Splunk, MaxPatrol, KUMA and R-Vision.
/ How it works
How DataSafe masks data and controls disclosure
This is how everyday work with a customer or employee profile changes after DataSafe is connected.
01
Opening a profile
Users work in their familiar CRM, HR system or internal service. The interface does not change.
02
Fields are hidden by default
Sensitive fields appear fully protected, not partly masked with asterisks.
03
Viewing follows a rule
Disclosure depends on role, work scenario and request context. The action is explicit and signed by the user.
04
The action is logged
The request enters the log. If it differs from the role profile, a signal is sent to existing security tools.
/ Why implement DataSafe
Personal-data protection results for business, security and the team
These are not features but measurable effects for commercial, security and operational teams after DataSafe goes live.
01 / 04
4 resultsProtection and compliance
01
Lower risk of a customer-database leak
Even if the database is compromised, the attacker receives anonymised values. Sensitive fields are stored separately and available only through DataSafe.
02
A protected layer in production
A separate personal-data access environment works within existing production systems, with a role model, visibility policy and access log.
03
Data management in test environments
Anonymised sets replace real records in test, demo and development. Compliance with Government Decree No. 1154 of 30 May 2025—without a separate project.
04
Making a breach worthless, even if it succeeds
An attacker who enters the infrastructure does not obtain real personal data. The value of the data falls to zero.
02 / 04
3 resultsInvestigation and control
05
Support for incident investigations
A complete access log linked to role, scenario and time. During an incident, evidence is ready in minutes instead of weeks of forensics.
06
Automated security control
Rules, behavioural analytics and dashboards. Security receives signals in existing SIEM/SOAR tools, not a separate product to administer.
07
Greater employee accountability
Every data disclosure is an explicit action with a signature and context. Transparency removes hidden access without putting pressure on the team.
03 / 04
1 resultTrust and reputation
08
Customer trust in personal-data processing
The ability to clearly tell a customer, regulator or auditor who accessed personal data, when and on what basis.
04 / 04
3 resultsOperations and resilience
09
No 24/7 employee required
Control works automatically through rules, logs and signals, without a dedicated around-the-clock monitoring service or night shifts.
10
Support for testing in test environments
Anonymised sets preserve statistical properties, formats and relationships. Testing, demonstrations and training remain realistic without restrictions.
11
Does not disrupt existing business processes
Connects to existing CRM and HR systems through targeted integrations, without rewriting products or retraining the team.
/ Where and how to use it
Where personal-data protection works without stopping processes
Start with one environment where risk is high and the effect is easy to show to leadership and security. Then scale to the other systems.
01
Sales and customer service
Unnecessary viewing of customer profiles in CRM is reduced without slowing service. Managers work as before but see only what the scenario requires.
02
HR and internal services
Employee, candidate, contractor and internal-user data is protected in HR systems and self-service portals.
03
Support and contractors
Access becomes targeted, while every data request is traceable and auditable. External teams do not receive excessive rights.
04
Development and testing
Anonymised data is used instead of real records in test, training and demo environments. Contractors work with the structure, not with PII.
How the first stage starts
- 01
Choose an environment
Sales, HR, support or a test environment—where the risk of excessive access is highest and the effect is easiest to demonstrate.
- 02
Define fields and roles
Decide which fields are fully hidden and who may disclose them in each scenario and on what basis.
- 03
Connect the system to DataSafe
Configure writing, reading and hiding of values in the selected environment, without rewriting CRM or replacing product forms.
- 04
Connect security to the process
Configure the access log, control rules and dashboards for the security team. Send signals to existing SIEM/SOAR.
/ Architecture and technology stack
How DataSafe fits into the corporate data-protection environment
DataSafe works inside the company environment and exchanges data only between corporate systems. The solution has no direct access from the internet.
01
Internal environment without a public perimeter
Works with CRM, HR and integration systems through targeted connections. Calls run between systems and are not exposed externally.
02
Encrypted connections with signature verification
All interactions use protected internal channels with signature verification and time-window control.
03
Event log without personal-data records
The fact and context of access are recorded, but personal data itself never enters the log. The log does not become a second copy of the database.
04
Signals to existing security tools
Sends events to Splunk, MaxPatrol, KUMA, R-Vision and other SIEM/SOAR tools in standard formats, without a separate monitoring console.
05
Anonymisation for test and development
Test, training and demo environments receive anonymised data sets instead of real values, in accordance with Government Decree No. 1154.
Image is loading
/ Scenarios
What the solution closes in the working environment
Each scenario is a working environment where the solution delivers measurable value.
— Scenario 01 —
Sensitive-field catalogue
One model of what counts as personal data in CRM, HR and integration systems, with a managed visibility policy.
— Scenario 02 —
Hiding data from employees
Values are masked based on role, scenario and environment—without changing storage methods or rewriting the product.
— Scenario 03 —
Anonymised storage
A separate data environment for analytics and testing. It is not coupled with production systems and never returns real personal data.
— Scenario 04 —
Breaking the individual/company link
Separating people from identifiers in technical environments without losing business logic or process relationships.
— Scenario 05 —
Audit log and behavioural analytics
An end-to-end history of access to sensitive data, with anomaly detection and integration into existing SIEM/SOAR.
/ Related materials
What often goes together
1 material
Ready to see DataSafe — Personal Data Protection on your data?
Thirty minutes with an engineer, no slides and no NDA for the demo.