1. Purpose of this policy
The Provider, TOO “Gamechanger”, attaches great importance to the security of its websites, the 4B software and its IT systems.
Despite all safeguards, no IT system can be considered completely free of vulnerabilities.
If you discover a possible vulnerability in our systems as a security researcher, customer, user or in any other capacity, we ask you to report it to us responsibly and in a coordinated manner (coordinated vulnerability disclosure) so that we can investigate and fix the problem.
This policy sets out:
- where to send reports;
- what information is helpful;
- which actions are permitted;
- which actions are not permitted;
- how we handle reports;
- how publication is coordinated;
- which legal framework we assure.
2. Contact
3. Scope
This policy applies to publicly accessible systems operated by the Provider, in particular:
- systems4b.de and its subdomains;
- the 4B catalog;
- the 4B SaaS platform, exclusively using your own test or customer accounts;
- public application programming interfaces (API) of 4B;
- other public services and applications of the Provider, insofar as no separate security policy applies to them.
This policy does not constitute general permission to test any of the Provider’s systems, infrastructure, servers or products.
If a system is not expressly mentioned or it is unclear whether it is operated by the Provider, please obtain our consent before any active testing.
4. Customer systems and data
This policy does not permit testing of our customers’ systems, tenants or data.
Even if a system:
- is based on 4B;
- is developed or maintained by the Provider;
- bears the name 4B or the name of the Provider;
- is technically connected to the Provider’s infrastructure,
it may belong to another company or contain data of a customer who is the controller under data protection law.
Testing such systems requires the consent of the respective owner. If you have come across a possible vulnerability by chance, you can report it to us; where possible, we will forward the information to the responsible owner.
5. Reportable vulnerabilities
In particular, you can report possible:
- flaws in access control;
- insecure direct object references (IDOR);
- authorisation flaws, in particular access to data of other tenants;
- authentication bypass;
- SQL injection;
- command injection;
- cross-site scripting (XSS);
- server-side request forgery (SSRF);
- cross-site request forgery (CSRF);
- insecure file uploads;
- security-relevant misconfigurations;
- exposed secrets and API keys;
- insecure interfaces;
- flaws in session management;
- disclosure of personal or confidential data;
- misconfigurations of the cloud infrastructure;
- vulnerable dependencies;
- flaws in business logic;
- privilege escalation;
- critical vulnerabilities in AI features;
- disclosure of the internal context of AI systems;
- access by AI features to data of other users;
- other problems that can actually affect the confidentiality, integrity or availability of a system.
6. How to report a vulnerability
Please send your report by email to: developers@gamechanger.kz
The contact details are also available in machine-readable form in the file /.well-known/security.txt in accordance with RFC 9116 on our website. On request, we will provide a way to submit reports in encrypted form.
The following is particularly helpful:
- URL or name of the affected system;
- type of the suspected vulnerability;
- a short description;
- steps to reproduce;
- expected behaviour;
- actual behaviour;
- possible impact;
- date and time of discovery;
- browser, device or other technical context, where relevant;
- a minimal proof of the vulnerability.
If necessary, you can attach:
- screenshots;
- a harmless proof of concept;
- HTTP requests;
- excerpts from server responses;
- logs without unnecessary personal data;
- a video of the reproduction.
Reports can be made in German or English.
7. Minimal intrusion
Please demonstrate a possible vulnerability with as little intrusion as possible.
If it is sufficient for the proof to:
- establish that access is possible;
- access a single test object of your own;
- show that an action can be performed without actually performing it,
please do not continue the exploitation.
The aim of a good-faith investigation is to demonstrate the problem – not to prove how large the possible damage would be.
8. Personal data
If you happen to come across personal data of other people during your investigation: Stop the investigation at this point.
Please:
- do not download any datasets;
- do not copy any databases;
- do not make any further requests for other people’s records;
- do not pass on any information found;
- do not publish screenshots containing identifiable personal data.
Record only the minimum strictly necessary for the proof, report the finding without delay and delete any copies of personal data once the coordination with us has been completed.
9. Trade secrets and confidential information
If a vulnerability has accidentally disclosed any of the following:
- source code;
- contracts;
- trade secrets within the meaning of the German Trade Secrets Act (GeschGehG);
- keys;
- internal documents;
- configurations;
- customer data;
- non-public repositories,
please stop gathering further information as soon as the problem has been sufficiently demonstrated.
Information obtained may only be used for the report to us and must be treated confidentially.
10. Access credentials and secrets
If you find:
- an API key;
- a password;
- a token;
- a private key;
- other access credentials;
- secrets of the cloud infrastructure,
please do not use them to investigate further systems.
In your report, it is sufficient to state:
- where the secret was found;
- which type of system it probably belongs to;
- minimal indications that it may be valid.
We will handle revocation, rotation and internal review ourselves.
11. Permitted actions
Under this policy, reasonable actions with minimal intrusion are permitted, in particular:
- evaluating publicly accessible information;
- analysing client-side components of the website;
- tests using your own data, your own accounts and your own sessions;
- careful verification of a suspected flaw;
- creating a minimal proof of concept;
- a limited number of requests required for reproduction;
- analysing public HTTP headers and configurations;
- reporting a problem discovered by chance.
Each individual case is assessed taking into account the nature of the actions and their effects.
12. Prohibited actions
Without separate written consent, the following are prohibited in particular:
- denial-of-service attacks (DoS/DDoS);
- generating significant load;
- mass scanning of the infrastructure;
- brute-force attacks;
- guessing other people’s passwords;
- credential stuffing;
- accessing other people’s accounts or tenants;
- modifying or deleting other people’s data;
- setting up persistent access;
- installing backdoors;
- introducing malware;
- encrypting data;
- ransomware scenarios;
- social engineering;
- phishing against employees;
- calling employees to obtain confidential information;
- unauthorised physical access, for example to offices or data centres;
- extracting databases;
- downloading datasets containing personal data;
- using secrets found to access other systems;
- modifying production data;
- disrupting other users;
- selling or passing on information about the vulnerability to third parties;
- demanding payment under threat of publication or exploitation.
13. Automated scans
Extensive automated scans are not permitted under this policy.
Even common scanners can:
- generate significant load;
- trigger lockout mechanisms;
- impair unstable features;
- cause numerous false alarms in security monitoring.
Intensive automated testing requires prior coordination with us.
14. Authorisation testing
If a suspected vulnerability concerns access to objects of other users or tenants, please use:
- your own test accounts, if necessary two accounts of your own;
- test data;
- a minimal, harmless scenario.
If registration is not possible and the assumption cannot be verified safely, please report the possible problem without further exploitation.
15. Security of AI features
We also accept reports of possible vulnerabilities in AI features, for example:
- disclosure of other users’ data;
- access to internal documents without the corresponding authorisation;
- disclosure of secret keys;
- bypassing access restrictions on knowledge bases;
- obtaining confidential system information;
- using the AI as a route to unauthorised access;
- dangerous flaws in the connection of AI to external tools.
An unusual or incorrect AI answer alone does not constitute a security vulnerability.
16. Prompt injection
Investigations into prompt injection are only permitted to an extent that does not create any actual risk to data or infrastructure.
Prompt injection must not be used to:
- obtain other people’s personal data;
- obtain keys;
- bypass actual authorisation;
- perform unauthorised actions in external systems;
- modify or delete data.
If you have found a way to get the AI to reveal internal instructions without sensitive data becoming accessible, you can report this as a possible problem.
17. No exploitation of vulnerabilities
A discovered vulnerability must not be used:
- to gain economic advantages at the expense of users;
- for extortion or coercion;
- to use services free of charge by bypassing payment mechanisms;
- to modify other people’s data;
- to obtain competitive information;
- for attacks on third parties.
Responsible disclosure ends where exploitation for one’s own benefit or to the detriment of others begins.
18. Coordinated publication
Please do not publish technical details of a vulnerability immediately after discovering it.
Before publication, give us reasonable time to:
- confirm the problem;
- assess its impact;
- develop a fix;
- test and roll out the fix;
- inform affected persons and customers where necessary;
- fulfil legal obligations.
19. Coordinating publication
If you wish to publish information about a vulnerability, please agree the following with us in advance:
- the time of publication;
- the extent of technical details;
- the redaction of personal data;
- the redaction of infrastructure information;
- the status of the fix.
As a guideline, we aim to allow publication after the fix, but no later than 90 days after receipt of the report. If a fix is not possible within this period for understandable reasons, we will agree a reasonable later date with you.
We do not want to prevent good-faith disclosure permanently, but to avoid instructions for exploitation being published before users are protected.
20. Critical vulnerabilities
The following are considered potentially critical in particular:
- remote code execution;
- complete bypass of authorisation;
- mass disclosure of personal data;
- cross-tenant access to customer data;
- access to the production infrastructure;
- compromise of administrative interfaces;
- obtaining critical secrets;
- the ability to modify essential data;
- vulnerabilities that enable attacks on a large number of users.
In these cases, please state in the subject line: SECURITY – CRITICAL
21. Process after receipt of a report
After receiving a report, we will usually:
- record the report and confirm receipt;
- carry out an initial assessment;
- try to reproduce the problem;
- assess the risk;
- request further information if necessary;
- develop a fix;
- test and roll out the fix;
- take measures to handle a security incident where applicable;
- inform you of the outcome.
22. Response times
We aim for the following response times:
- acknowledgement of receipt: within three working days;
- initial assessment: within ten working days;
- updates on the processing status: at reasonable intervals, with priority for critical vulnerabilities.
The further duration depends in particular on:
- the severity of the problem;
- the effort required for reproduction;
- the system affected;
- the involvement of external providers;
- the type of fix required.
The times stated are target values; this policy does not create any entitlement to compliance with them.
23. Risk assessment
In our assessment, we take into account, among other things:
- exploitability;
- required privileges;
- the need for user interaction;
- the scope of accessible data;
- the possible damage;
- the number of affected users;
- the complexity of an attack;
- existing compensating controls;
- the possibility of remote exploitation.
Recognised methods such as the Common Vulnerability Scoring System (CVSS) may be used for the technical assessment. The final prioritisation also takes the specific operational context into account.
24. Duplicate reports
Several people may discover the same problem independently of each other.
If a report concerns a vulnerability already known to us, it may be classified as a duplicate. As a rule, the time of receipt of the first complete report is decisive.
This does not mean that your report was useless.
25. Findings without direct security relevance
Some reports are classified not as a vulnerability but as a hardening recommendation, for example:
- the absence of an optional security header;
- the disclosure of information that is already public;
- theoretical attacks without a practically feasible scenario;
- software versions without a proven, actually applicable vulnerability.
Such findings can also be useful and are reviewed by us.
26. No obligation to pay rewards
This policy does not establish a public bug bounty programme; there is no entitlement to remuneration or a reward.
If the Provider decides to give recognition or a reward in an individual case, this is done voluntarily, on a case-by-case basis or as part of a separately published programme.
Submitting a report does not create a contractual relationship or any payment obligation on the part of the Provider.
27. No commissioning of security tests
This policy is not an order or an agreement for:
- penetration tests;
- security audits;
- red team exercises;
- load and performance tests;
- comprehensive scans of the infrastructure.
Such services require a separate written agreement.
28. Acting in good faith
When assessing your actions, we take into account in particular whether you:
- acted with the aim of reporting a problem to us;
- limited the impact;
- stopped the investigation once there was sufficient proof;
- did not use other people’s data;
- did not cause any intentional damage;
- did not seek any unlawful advantages;
- reported the problem responsibly and confidentially;
- otherwise complied with this policy.
29. Legal framework and assurance (safe harbour)
If you investigate and report a vulnerability in good faith and in compliance with this policy, we assure you of the following within the scope of applicable law:
- we consider actions that this policy expressly describes as permitted to be authorised by us for our own systems;
- we will not assert any civil claims against you;
- we will not file a criminal complaint because of your investigation, in particular not for data espionage or interception of data (Sections 202a, 202b StGB) or data alteration or computer sabotage (Sections 303a, 303b StGB);
- we will not disclose your identity to third parties without your consent unless we are obliged to do so by law or by an authority.
Important limitations:
- This assurance does not exempt you from any criminal liability. Law enforcement authorities may prosecute offences that are prosecuted ex officio or in which they assume a special public interest, regardless of our position;
- we cannot grant authorisation for systems, data or rights of third parties, in particular not for systems of our customers or of hosting, cloud or AI providers; these third parties are not bound by this assurance;
- the assurance does not apply to actions that violate this policy or mandatory law.
If you are unsure whether a particular action is permitted, please pause and ask us first.
30. Confidentiality of the reporter
In your report, you can state:
- your name;
- a pseudonym;
- your organisation;
- a contact for follow-up questions.
Anonymous reporting is possible; however, follow-up questions and updates on the processing status are then not possible.
If you do not wish to be named in a publication about the fix, we will take this into account.
31. Acknowledgements
With your consent, we can publicly thank you for a responsible report, for example on an acknowledgements page.
The following may be published, for example:
- your name;
- a pseudonym;
- a profile;
- the name of your organisation.
We do not publish any personal data about you without your consent.
32. Data protection
We process personal data received in connection with a report in accordance with our privacy policy and the GDPR.
The legal basis is Art. 6 (1) (f) GDPR; our legitimate interest lies in ensuring the security of our systems and communicating with you. Where legal obligations exist, the legal basis is Art. 6 (1) (c) GDPR; a public acknowledgement is only made with your consent (Art. 6 (1) (a) GDPR).
The data is stored insofar as this is necessary for:
- the investigation;
- fixing the problem;
- ensuring security;
- preventing repeated incidents;
- fulfilling legal obligations.
It is then deleted, at the latest three years after the matter has been closed, unless longer retention obligations apply.
33. Personal data breaches
If a report indicates a possible personal data breach, we carry out a separate assessment and fulfil our legal obligations, in particular notification of the competent supervisory authority within 72 hours (Art. 33 GDPR) and, where applicable, notification of the data subjects (Art. 34 GDPR). If our customers’ data is affected, we inform them in accordance with the data processing provisions in the SaaS agreement.
Please do not inform potentially affected persons yourself without prior coordination with us, unless this creates no immediate danger to persons and the law does not require otherwise.
34. Third-party providers
Some problems may originate:
- in a third-party library;
- in a cloud service;
- with an AI provider;
- in an analytics system;
- in other external technology.
In this case, we may:
- take compensating measures;
- update the component;
- deactivate the feature;
- inform the respective provider;
- provide it with the necessary technical information.
In doing so, we only pass on personal data of the reporter or of users insofar as this is necessary.
35. Emergency contact
Please do not transmit critical secrets via unprotected channels if stating where they were found is sufficient to describe the problem.
36. Changes to this policy
We may adapt this policy, in particular in the event of:
- changes to the infrastructure;
- the introduction of new services;
- changes in the legal situation;
- developments in the field of AI;
- changes to information security processes;
- the launch of a bug bounty programme.
The version published on our website applies.