Smarter Testing, Safer AI
Last updated: 12 September 2025
Gates-AI (a subsidiary of Gates Digital Pte. Ltd., Singapore) provides independent AI assurance: testing, audits, risk assessments, benchmarking, MLOps pipeline testing, and certification services.
We perform services set out in an Order, Proposal, or SOW (collectively, “SOW”). Unless stated otherwise, services include: model functional testing, prompt/LLM testing, bias & fairness audits, compliance testing, stress & edge-case testing, safety & risk assessments, performance benchmarking, and MLOps pipeline testing. Certification (where applicable) follows our Gates-AI Certification Framework and published criteria.
You will: (a) provide timely access to systems, environments, datasets, and SMEs; (b) ensure you have lawful rights to share data and systems for testing/audits; (c) maintain backups and operational safeguards; (d) not request unlawful testing (e.g., unauthorized access).
We apply industry controls (least privilege, encryption, logging, change control, vulnerability management, incident response) consistent with ISO/IEC 27001 practices. You may request our current security summary and subprocessors list under NDA.
Unless expressly authorized in writing, we will not conduct destructive tests in production. You will provide approved windows, targets, and test data. Findings may include simulated prompt injection, adversarial inputs, stress loads, and bias tests.
Each party will protect the other’s Confidential Information using reasonable safeguards and use it solely to perform the SOW.
We perform services consistent with NIST AI RMF and ISO/IEC 42001 principles for AI risk/governance; and relevant privacy laws (GDPR/PDPA/CPRA) where applicable.
We warrant we’ll perform services with reasonable skill and care. We do not warrant that your systems will be error-free, invulnerable, or fully compliant after remediation—risk changes over time.
To the maximum extent permitted by law:
Either party may terminate for material breach uncured within 30 days. Upon termination, you pay for work performed. Confidentiality, IP, and liability limits survive.
We will not use your name or marks without permission, except to identify you as a client where you have publicly disclosed the relationship.
Each party complies with applicable export controls and sanctions.
Neither party is liable for delays beyond reasonable control (e.g., outages, disasters, war).
Unless agreed otherwise in the SOW: Singapore law governs; courts of Singapore have exclusive jurisdiction.
If there’s a conflict, the SOW controls over these T&Cs.
Gates AI make Artificial Intelligence (AI) reliable, fair, and secure. Our expert team delivers rigorous testing, ethical audits, and compliance checks to ensure AI systems work flawlessly and responsibly. From data validation to post-deployment monitoring, we help organizations deploy AI with confidence and trust.