Any design, implementation or configuration issue that substantially affects confidentiality or integrity is likely to be eligible for a reward. Common examples include:
- Remote code execution (RCE)
- Code injections (HTML, JS, SQL, PHP, etc)
- Cross-site scripting (XSS)
- Horizontal and vertical privilege escalation
- Broken authentication and session management
- Sensitive data exposure
The following issues are not considered relevant under the Programme:
- The absence of a security feature alone without demonstrated impact
- Disclosure of non-sensitive information, even in bulk
- Denial-of-Service attacks
- "Self" XSS without demonstration of further impact
- Missing Cookie flags or missing security-related HTTP headers without a demonstrated security impact
- Open redirects without demonstration of further impact
- Clickjacking
- Missing or incomplete SPF/DKIM/DMARC
- Social engineering attacks
- Reports from automated tools or scans without explanatory documentation
- Missing Rate limiting without further security impact
Moreover, in cases where we generally accept the security impact, we also exclude certain classes of vulnerabilities.
| Date | Vulnerability | Comment |
|---|---|---|
| Jan 2020 | Disclosure of unrestricted API keys | Disclosed API keys used for services like Google, Facebook, etc may lead to financial loss and/or denial of service if a quota is set and exhausted. These will no longer be considered as relevant for our Bug Bounty Programme. |
| Jun 2022 | XSS in legacy Browsers | Cross-site scripting (XSS) in legacy browsers such as Microsoft Internet Explorer will no longer be considered for our Bug Bounty Programme |
| Jul 2022 | CVE-2019-12616, CSRF in phpMyAdmin < 4.9.0 | Proof-of-concept requires user interaction, which violates our policy on social engineering. After weighing out the risk it was decided that this attack will no longer be considered as relevant for our Bug Bounty Programme. |
| Jul 2022 | CWE-489 Active Debug Code | Unless sensitive data is exposed, applications running with debug mode turned on do not constitute a security issue considered as relevant for our Bug Bounty Programme. |
| May 2023 | Takeover of expired DNS domains listed in inscope.txt. | Expired domains left in scope are a result of managing a broad programme, where the bug bounty team is not involved in the domain lifecycle of all participating entities. Unless a domain takeover is possible (without registering the domain itself), it will be assumed that expired domains are intentional. |
| Aug 2023 | Dangling DNS records without PoC | Reports on dangling DNS records will be rejected unless proof of successful (sub-)domain hijacking is provided. |
| Oct 2023 | Outdated Browser version as prerequisite | Reports relying on the usage of not up-to-date browsers will be rejected. |
| Jan 2024 | Harmless XSS payloads | XSS payloads without demonstration of a malicious impact will be disregarded. For example the injected code must affect the same-origin as the target application, or it should be able to retrieve sensitive information from the cookies/local storage. For details, see "Do NOT use alert(1) in XSS". |
| Jun 2024 | Leaked Credentials | Reports of leaked credentials e.g. originating from infostealers, password dumps, etc. are not eligible for bounties. |
| Dec 2024 | NTLM over HTTP Information Disclosure | Host information can be enumerated when NTLM over HTTP authentication is supported and exposed. This is considered as disclosure of non-sensitive information (w.r.t. data protection laws) and will be rejected unless further impact is demonstrated. |
Most non-critical patches are applied during standard maintenance windows. Therefore, 1-day vulnerabilities are accepted four weeks after disclosure at the earliest (see exceptions below).
Special cases where information on 1-day vulnerabilities is explicitly desired are listed below. For any clarifications regarding this list, please contact the Bug Bounty team.
| Date | Vulnerability | Comment |
|---|---|---|
| Dec 2021 | CVE-2021-44228 (Log4Shell) |