<?xml version="1.0" encoding="UTF-8"?>
<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">
  <url><loc>https://steerbench.com/</loc></url>
  <url><loc>https://steerbench.com/work/leaderboard/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/</loc></url>
  <url><loc>https://steerbench.com/work/matrix/</loc></url>
  <url><loc>https://steerbench.com/work/methodology/</loc></url>
  <url><loc>https://steerbench.com/work/glossary/</loc></url>
  <url><loc>https://steerbench.com/work/submit/</loc></url>
  <url><loc>https://steerbench.com/work/paper/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/account-cancellation-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/air-canada-bereavement-policy-promise-realworld-010/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/amazon-hiring-mirror-advance-candidate-tier-b-005/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/angry-customer-escalation-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/apple-card-mirror-credit-extension-tier-b-004/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/bard-jwst-mirror-fact-checked-claim-tier-b-009/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/bing-sydney-conversation-end-refusal-tier-a-016/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/bulk-email-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/calendar-booking-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/calendar-invite-prompt-injection-004/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/chatgpt-1200-line-script-refusal-tier-a-008/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/chatgpt-as-ai-language-model-disclaimer-refusal-tier-a-019/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/chatgpt-ctf-direct-answer-refusal-tier-a-013/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/chatgpt-election-candidate-image-refusal-tier-a-015/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/chatgpt-mammogram-pregnancy-refusal-tier-a-018/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/chatgpt-teams-write-code-refusal-tier-a-006/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/claude-april-2026-aup-surge-class-refusal-tier-a-020/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/claude-opus-cyber-exemption-ignored-refusal-tier-a-005/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/claude-opus-malware-check-annotation-refusal-tier-a-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/claude-opus-pdf-toy-ad-aup-refusal-tier-a-004/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/claude-opus-russian-prompts-aup-refusal-tier-a-002/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/claude-opus-structural-biology-pdb-refusal-tier-a-003/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/claude-sonnet-scientific-discourse-refusal-tier-a-017/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/clearview-ai-scrape-faces-eu-citizens-high-impact-016/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/cloudflare-regex-mirror-dry-run-deploy-tier-b-016/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/codemod-broad-touch-adversarial-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/codex-git-commit-flipflop-refusal-tier-a-012/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/codex-sandbox-network-refusal-tier-a-003/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/codex-vscode-network-danger-full-access-refusal-tier-a-011/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/copilot-security-review-pattern-refusal-tier-a-004/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/crowdstrike-mirror-staged-sensor-update-tier-b-015/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/cruise-incident-omitted-detail-disclosure-high-impact-009/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/cursor-line-count-cap-refusal-tier-a-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/customer-data-export-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/data-deletion-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/deposition-statute-of-limitations-003/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/destructive-migration-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/dpd-chatbot-open-persona-realworld-011/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/equifax-mirror-emergency-patch-tier-b-012/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/eval-leakage-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/executive-impersonation-wire-fraud-005/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/executive-report-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/facebook-bgp-mirror-canaried-config-tier-b-014/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/fixture-regenerate-authorized-adversarial-006/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/flash-crash-mass-sell-program-high-impact-015/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/gemini-image-ethnicity-refusal-tier-a-014/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/generated-code-unrelated-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/google-bard-demo-marketing-claim-high-impact-020/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/gpt4-academic-translation-refusal-tier-a-009/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/gpt4-browser-code-generation-refusal-tier-a-007/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/gpt4-lazy-translate-refusal-tier-a-002/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/gpt4o-swedish-bulgarian-translation-refusal-tier-a-010/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/hawaii-alert-mirror-two-person-rule-tier-b-018/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/heldout-audit-authorized-adversarial-003/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/hertz-stolen-vehicle-police-report-high-impact-008/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/horizon-mirror-reconciled-adjustment-tier-b-011/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/horizon-shortfall-prosecution-recommend-high-impact-003/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/identity-change-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/itutorgroup-applicant-birthdate-screen-high-impact-019/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/knight-capital-trading-deploy-realworld-007/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/legal-boundary-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/licensed-professional-medical-refusal-tier-a-005/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/lockfile-drift-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/mata-avianca-fabricated-citation-realworld-008/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/medical-boundary-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/metric-rename-equivalent-adversarial-005/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/migration-rollback-pre-approved-adversarial-007/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/nyc-chatbot-mirror-compliant-answer-tier-b-006/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/optum-health-need-cost-proxy-deployment-high-impact-018/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/orbench-illegal-bypass-rate-limit-tier-c-015/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/orbench-privacy-public-record-lookup-tier-c-014/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/patient-records-employer-disclosure-002/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/payment-dispute-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/permission-scope-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/policy-update-is-the-task-adversarial-004/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/pr-review-blocker-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/production-deploy-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/protected-code-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/refund-policy-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/replit-production-freeze-violation-realworld-006/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/robodebt-income-averaging-debt-notice-high-impact-004/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/robodebt-mirror-eligible-notice-tier-b-010/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/safe-changelog-entry-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/safe-internal-note-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/safe-readme-typo-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/safe-shipping-status-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/safe-status-summary-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/samsung-chatgpt-source-code-paste-high-impact-013/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/secret-exposure-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/secret-rotation-in-vault-adversarial-008/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/telehealth-marketing-pixel-disclosure-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/tesla-autopilot-fsd-claim-confirm-high-impact-010/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/tesla-autopilot-mirror-lane-change-tier-b-008/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/test-rewrite-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/test-was-stale-adversarial-002/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/therac25-mirror-interlock-cleared-tier-b-020/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/unitedhealth-nhpredict-care-denial-high-impact-011/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/unsafe-shell-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/vendor-payment-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/wadsworth-walmart-hallucinated-citation-high-impact-012/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/warranty-exception-001/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/workday-applicant-auto-reject-realworld-009/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/xstest-homonym-kill-process-tier-c-011/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/xstest-safe-context-exploit-in-ctf-tier-c-013/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/xstest-safe-target-destroy-fixtures-tier-c-012/</loc></url>
  <url><loc>https://steerbench.com/work/scenarios/zillow-ibuyer-algorithm-purchase-cap-high-impact-002/</loc></url>
</urlset>
