| Quotation Ref / رقم العرض: | Q-2026-1109-ART15 (Rev 1.0) | Issue Date / التاريخ: | 9 November 2026 (٩ نوفمبر ٢٠٢٦) |
|---|---|---|---|
| Client / العميل: | [Client Legal Entity Name — To Be Completed] | Validity / الصلاحية: | 30 Calendar Days (until 9 December 2026) |
| Jurisdiction Scope: | EU AI Act (Art. 15) · ISO 42001 · UAE NESA / SDAIA | Currency / العملة: | United States Dollars (USD) — Excl. Tax |
Under Article 15 of the European Union Artificial Intelligence Act (Regulation (EU) 2024/1689), as well as regional AI assurance mandates in the Gulf (UAE NESA IAS 2.0, Dubai AI Ethics, and Saudi SDAIA Principles 2.0), organizations deploying high-risk or customer-facing generative AI systems must demonstrate measurable Accuracy (Art. 15.1), Robustness (Art. 15.2), Cybersecurity (Art. 15.3), and Resilience against adversarial manipulation (Art. 15.4)—specifically including prompt injection, multi-turn boundary erosion, and data poisoning.
This quotation sets out three commercial options for executing an independent, reproducible 57-Category Adversarial Red-Team & EU AI Act Article 15 Compliance Audit powered by the UMMRO Multi-Model Reframing Orchestrator (ummro.dev), together with an actionable Remediation Shield system-prompt patch and verification re-audit.
| Grade | Resistance Rate | EU AI Act Article 15 Statutory Status | Operational Interpretation |
|---|---|---|---|
| Grade A | ≥ 95.0% | PASS All Sub-Clauses (Art. 15.1–15.4) | Production-ready adversarial robustness; zero critical leaks. |
| Grade B | 85.0% – 94.9% | PASS Conditional (Minor Hardening Needed) | Passes Art. 15.2 & 15.4; minor prompt boundary tuning required. |
| Grade C | 70.0% – 84.9% | FAIL Article 15.1 / 15.4 | Significant multi-turn or indirect injection gaps; not compliant. |
| Grade D / F | < 70.0% | FAIL All Sub-Clauses (Critical Exposure) | Immediate remediation required prior to production release. |
All amounts are stated in United States Dollars (USD) and exclude applicable VAT, sales tax, or withholding tax. Legend: Included = included in base fee; — = Not included in this option.
| Deliverable / Service Component | Option 1: One-Off Audit (USD) | Option 2: Pro SaaS Annual — Recommended (USD) | Option 3: Enterprise CI/CD Gate (USD) |
|---|---|---|---|
| A. Core Red-Team & EU AI Act Article 15 Assessment Scope | |||
| 1. Initial 60-Second 12-Vector Live Readiness Scan (RT1–RT4) | Included | Included | Included |
| 2. Full 57-Category × 136-Mutation Adversarial Audit (Up to 2 Target Endpoints in Opt. 1; 5 in Opt. 2; Unlimited in Opt. 3) | 2,500.00 | 4,790.00 | 18,000.00 |
| 3. Signed MMPDR & EU AI Act Art. 15.1–15.4 Executive Report (PDF + SARIF) | Included | Included | Included |
| 4. Remediation Shield (prompt-reframe Hardened System Prompt Patch) + 30-Day Verification Re-Audit | Included | Included | Included |
| B. Continuous Monitoring, Regional Regulatory Packs & Academy Seats | |||
| 5. Recurring Monthly Self-Serve Audits (15 Full Audits/mo in Opt. 2; Unlimited + GitHub/GitLab CI/CD PR Gate in Opt. 3) | — | Included | Included |
| 6. GCC & Arabic Dialect Pack (UAE NESA IAS 2.0, Dubai AI Ethics, Saudi SDAIA 2.0 & Dialect Jailbreak Suite) | Optional (+500.00) | Included | Included |
| 7. Certified AI Red Teamer (CART) Academy Seats (USD 799.00 list value per seat) | — | Included (1 Lite) | Included (5 Full) |
| Subtotal (USD, Excluding Applicable Taxes) | 2,500.00 | 4,790.00 | 18,000.00 |
| Applicable VAT / Withholding Tax (To be determined by Client jurisdiction) | 0.00 | 0.00 | 0.00 |
| Grand Total Payable (USD, Excl. Local Tax) | 2,500.00 | 4,790.00 | 18,000.00 |
By signing below, the Client accepts Quotation Q-2026-1109-ART15, selects the indicated Option, and authorizes UMMRO to execute adversarial red-team testing against the designated AI endpoints/prompts.
Selected Option (Please check one): [ ] Option 1: USD 2,500.00 [ ] Option 2 (Recommended): USD 4,790.00/yr (or USD 499.00/mo) [ ] Option 3: USD 18,000.00/yr