← back
How to Build Trustworthy AI — Allie Howe
Takeaway
Trustworthy AI requires shifting right with runtime guardrails alongside build-time scanning and red-teaming because non-determinism means CI/CD alone can't catch failures.
Summary
- Allie Howe (vCISO, Growth Cyber) catalogs trust failures: 2023 Chevy Tahoe $1 chatbot, 2024 Slack prompt-injection data leak from private channels, Fortnite Darth Vader NPC with offensive speech.
- Legal reality: a radio host's defamation suit vs OpenAI was dismissed on the basis that ChatGPT can err and users must verify — so deployers bear liability and reputational risk.
- Trustworthy AI = AI security (outside-in attacks) + AI safety (model harm to world); needs collaboration across product, engineering, and security teams.
- Traditional DevSecOps doesn't fit AI workflows (Databricks, notebooks) — shift right matters more: build (model scanning, AI-BOM, ML provenance), test (red teaming), runtime (guardrails) all needed.
ai-securitytrustworthy-aired-teaming
Original description
Trust is a multifaceted outcome that results when product and engineering teams work together to build AI that is aligned, explainable, and secure. Learn strategies for how to build trustworthy AI and why trust is paramount for AI systems. Trustworthy AI = AI Security + AI Safety Learn about the differences between AI Security and AI Safety and how the three focus areas of MLSecOps + AI Red Teaming + AI Runtime Security can help you achieve both and ultimately build Trustworthy AI. Trustworthy AI Issues in the news: https://x.com/syddiitwt/status/1923427722241487297 https://fingfx.thomsonreuters.com/gfx/legaldocs/egvblxokkvq/Walters%20v%20OpenAI%20-%20order.pdf?ref=claritasgrc.ai MLSecOps Resources Modelscan https://github.com/protectai/modelscan Community: mlsecops.com AI Red Teaming Resources: https://azure.github.io/PyRIT/ https://ashy-coast-00aeb501e.6.azurestaticapps.net/MS_AIRT_Lessons_eBook.pdf AI Runtime Security Resources: https://www.pillar.security/solutions#ai-detection https://noma.security/ Showcasing Trustworthy AI to Customers/Prospects https://www.vanta.com/collection/trust/what-is-a-trust-center