Answer: The program is scoped to jailbreaks on Claude's Constitutional Classifiers, primarily universal jailbreaks that reveal harmful information across a wide range of queries and detailed jailbreaks revealing highly specific harmful information related to biological threats. Bounties are up to $35,000 per novel, universal jailbreak identified. Status: Confirmed Entry: tool/claude-apps/permissions-and-security/bug-bounty-scope Revision: 2 Digest: sha256:3797dc3cf0c8a449 (changes whenever the content changes) Timeline: Created: 2026-11-13T02:57:51Z Last changed: 2026-11-15T12:00:09Z (Maximum bounty: — → $35,000) Last verified: 2026-11-20T12:01:30Z Next check: 2026-11-21T12:01:30Z Check interval: Daily Expected change: No fixed schedule Data: Field | Value | Unit | Source | As of Scope | jailbreaks on our Constitutional Classifiers | | [1] | 2026-11-20 Jailbreak type | universal | | [1] | 2026-11-20 Maximum bounty | $35,000 | USD | [1] | 2026-11-20 Sources: [1] Anthropic. "Model Safety Bug Bounty Program | Claude Help Center". https://support.claude.com/en/articles/12119250-model-safety-bug-bounty-program Role: primary. Accessed: 2026-11-20T12:01:30Z Quote: "This program is scoped to jailbreaks on our Constitutional Classifiers." (Supports: Scope) Quote: "discovering jailbreaks that are universal" (Supports: Jailbreak type) Quote: "We will pay up to $35,000 per novel, universal jailbreak identified." (Supports: Maximum bounty) Snapshot SHA-256: e61afbff2fd07880e9b51d06d96509ad1d149326c3bcc6fd6b25704dd8da5963 (captured 2026-11-14T11:57:38Z) Verification: Method: Automated extraction from [1] with deepseek-flash (JSON output, thinking disabled). Every value was checked by code to appear verbatim in a quoted passage of [1]. On this check the text of [1] was unchanged since 2026-11-14T11:57:38Z, so the values verified then were kept. Cross-check: not-available at 2026-11-20T12:01:30Z. No independent second source is configured for this entry Last successful check: 2026-11-20T12:01:30Z Last attempt: 2026-11-20T12:01:30Z Notes and limitations: Extracted automatically from a single official source (section "Program Scope"). Related: https://latestn.com/tool/claude-apps/permissions-and-security/build-child-safety-program How do I build a child safety program for a product using Claude? https://latestn.com/tool/claude-apps/permissions-and-security/change-project-public-to-private How do I change a Claude project from public to private? https://latestn.com/tool/claude-apps/permissions-and-security/child-safety-usage-policy-obligations What child safety obligations do developers have under Anthropic's Usage Policy when building with Claude? https://latestn.com/tool/claude-apps/permissions-and-security/approve-salesforce-changes How do I review and approve changes to Salesforce in Claude? https://latestn.com/tool/claude-apps/permissions-and-security/audit-log-permissions Who can access and export Claude audit logs? https://latestn.com/tool/claude-apps/permissions-and-security/bug-bounty-apply How do I apply to the Claude Model Safety Bug Bounty Program? https://latestn.com/tool/ All entries in AI tools Change log: Time | Event | Change | Source 2026-11-15T12:00:09Z | changed | Maximum bounty: — → $35,000 | [1] 2026-11-15T12:00:09Z | changed | Jailbreak type: — → universal | [1] 2026-11-15T12:00:09Z | changed | Scope: — → jailbreaks on our Constitutional Classifiers | [1] 2026-11-15T12:00:09Z | changed | maximum bounty: $35,000 → — | [1] 2026-11-15T12:00:09Z | changed | jailbreak type: universal → — | [1] 2026-11-15T12:00:09Z | changed | scope: jailbreaks on our Constitutional Classifiers → — | [1] 2026-11-13T02:57:51Z | added | Entry created | — Changes for this entry as a feed: https://latestn.com/tool/claude-apps/permissions-and-security/bug-bounty-scope/latest Challenge this entry: Open challenges: None How: If a value here is wrong or out of date, send what is wrong and a link to evidence. Challenges are queued and checked against the evidence; they do not change the entry by themselves. If the entry changes as a result, the change log says so. Form: https://latestn.com/dispute/tool/claude-apps/permissions-and-security/bug-bounty-scope POST: https://latestn.com/dispute/tool/claude-apps/permissions-and-security/bug-bounty-scope with claim, evidence_url, and optionally field and contact (form-encoded or JSON) GET: https://latestn.com/dispute/tool/claude-apps/permissions-and-security/bug-bounty-scope?claim={what is wrong}&evidence_url={link} How to cite: LatestN. (2026). What is in scope for the Claude Model Safety Bug Bounty Program? Revision 2, last verified 2026-11-20T12:01:30Z. https://latestn.com/tool/claude-apps/permissions-and-security/bug-bounty-scope (accessed 2026-11-21T10:04:12Z). @misc{LatestN_tool_claude_apps_permissions_and_security_bug_bounty_scope_r2, title = {What is in scope for the Claude Model Safety Bug Bounty Program?}, author = {{LatestN}}, year = {2026}, howpublished = {\url{https://latestn.com/tool/claude-apps/permissions-and-security/bug-bounty-scope}}, note = {Revision 2; digest sha256:3797dc3cf0c8a449; last verified 2026-11-20T12:01:30Z; accessed 2026-11-21T10:04:12Z} } Reuse: License: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/) Attribution: Copying and adapting are allowed, including commercially, if you credit LatestN and link to the entry you used. Each entry ends with a ready-made citation. URL: https://latestn.com/tool/claude-apps/permissions-and-security/bug-bounty-scope