Responsible disclosure
Found a security issue in Jubi? We'd rather hear it from you than read about it. This page is what's in scope, how to send it to us, and what we'll do for researchers acting in good faith.
How to report
Email security@jubi.my. Include:
- A clear description of the issue and its security impact.
- Steps to reproduce, ideally with proof-of-concept code or a video demonstration.
- Your contact details so we can follow up. Reports may be submitted anonymously, but we cannot extend acknowledgement or recognition to anonymous reporters.
- Whether you intend to publish the finding, and a proposed disclosure date.
A signed PGP key for encrypted reports is on the roadmap. Until then, please redact any customer data in your report and refer to records by an opaque identifier.
Scope
The following are in scope:
- Production properties operated by Jubi:
jubi.myand its subdomains, the customer console, and the Jubi-operated platform endpoints. - The Jubi mobile and web client applications.
- Vulnerabilities affecting authentication, authorisation, tenant isolation, the audit log, or AI safety boundaries (prompt injection, output validation bypass, Atlas grounding bypass).
The following are out of scope:
- Third-party systems Jubi depends on but does not operate (model providers, hyperscaler infrastructure, customer-side IdPs and warehouses). Report those to the operator of the affected system.
- Social engineering of Jubi staff, customers, or contractors.
- Physical attacks on Jubi or its data centres.
- Denial-of-service or volumetric attacks.
- Findings that require a privileged position the attacker would not realistically have (compromised employee endpoint, root on the customer's machine).
- Issues already known to us, already publicly disclosed, or already reported by another researcher.
- Best-practice or hardening recommendations that do not constitute a vulnerability (e.g. missing security headers without a demonstrable impact).
- Automated scanner output without manual validation and demonstrated impact.
Safe harbour
If you research in good faith and within this policy, Jubi will:
- Treat the research as authorised under applicable computer-misuse laws to the extent we are able to do so under the law;
- Not pursue civil action against you for activities that are within scope and consistent with this policy; and
- Work with you on coordinated disclosure.
Safe harbour does not extend to anyone subject to U.S. or other applicable sanctions, or located in a jurisdiction sanctioned by the Government of Malaysia or our payment processors.
Coordinated disclosure
- We aim to acknowledge a valid report promptly, triage soon after, and keep you reasonably updated as we work on a fix.
- Standard coordinated disclosure window is 90 days from the date we acknowledge the report. We may request an extension if the fix requires significant engineering work, vendor coordination, or carries customer-side complexity. We will explain the reason if we do.
- We may request that you delay public disclosure until customers have had a reasonable opportunity to patch or update.
- If we and the reporter cannot agree on a disclosure timeline, we will discuss in good faith and take reasonable steps to protect customers in the interim.
Recognition
For valid reports of new, in-scope vulnerabilities we will, with your consent, acknowledge you in our security acknowledgements (planned for GA). Where local law permits and we are otherwise able, we may extend a recognition reward; this is at our sole discretion and is not a contractual entitlement.
We do not currently operate a paid bug bounty programme. If we launch one, the terms (eligibility, payout schedule, exclusions) will be published separately.
What we ask researchers to avoid
- Do not access, modify, or save customer data beyond what is strictly necessary to demonstrate impact. If you encounter customer data, stop and report.
- Do not run scans or tests that materially affect availability for other users.
- Do not perform testing during periods we have asked you to pause (e.g. during an active customer incident).
- Do not use findings for any purpose other than coordinated disclosure to Jubi.
- Do not publicly disclose details before the agreed disclosure date.
Customer-side red teaming
Customers may, under their engagement, conduct authorised security testing of their own tenant. This is not a vulnerability programme; it is a contracted activity. Email security@jubi.my in advance with the planned scope, window, and source IP ranges so we can avoid blocking legitimate test traffic and so the activity is correctly attributed in our logs.
Reservations
Jubi reserves the right to update or withdraw this policy, modify scope, decline to triage reports outside scope, decline recognition, and take any action it considers appropriate to protect its customers, its platform, and other researchers. Nothing in this policy creates a binding contractual obligation; the safe-harbour commitments above are extended as a matter of programme conduct.
Reports: security@jubi.my · Machine-readable contact: security.txt