David Campbell
David Campbell is the Head of AI Security Research at Scale AI with nearly two decades of experience scaling infrastructure, security, and AI systems across Silicon Valley’s leading technology companies. He is a Founding Member of the OWASP AI Vulnerability Scoring System (AIVSS) Project and a Founding Consortium Member of AIUC-1, the world’s first AI agent certification standard.
David is a leading voice in operational AI risk and has been recognized by the White House, United States Congress, and global government bodies for his pioneering work in AI red teaming and responsible AI deployment.
At Scale AI, David leads AI Security Engineering and spearheads the company’s responsible AI initiatives. He pioneered “Discovery”, one of the first large-scale AI red teaming platforms deployed across government agencies and Fortune 100 companies to probe and harden generative models against real-world threats. Under his leadership, the platform has become instrumental in evaluating AI systems for safety, reliability, and security vulnerabilities before deployment. Read From AI Safety to Enterprise Scale: Meet the First Speakers of ODSC AI East 2026.
David’s work has received significant recognition from government institutions worldwide. In August 2023, he discussed responsible AI deployment at the White House as part of the Biden Administration’s AI safety initiatives. That same month, he led the architecture of Scale AI’s evaluation platform for the DEF CON 31 AI Village, the largest-ever public generative AI red teaming event. The exercise brought together 2,244 participants to test eight large language models from Anthropic, Cohere, Google, Hugging Face, Meta, Nvidia, OpenAI, and Stability AI. Read It’s generative AI models vs. hackers at DEF CON’s AI Village.
In January 2024, David briefed members of the U.S. Congress AI Caucus on enterprise AI security challenges. He presented on AI security implications for allied nations at NATO Headquarters in Brussels in May 2024. In January 2025, he led an AI Red Teaming workshop for UK lawmakers at UK Parliament in London. His international engagements have extended to leading workshops at the Qatar National Cyber Security Agency’s 4th Cyber Governance Conference in September 2025 and the Seoul Forum on AI Safety & Security for international security leaders in October 2025.
David is a Founding Member of the OWASP AI Vulnerability Scoring System (AIVSS) Project, developing a standardized framework for evaluating AI security risks. AIVSS extends the Common Vulnerability Scoring System (CVSS) for AI-specific vulnerabilities, including model manipulation, data poisoning, and agent misalignment. The project aims to release version 1.0 by the RSA Conference in March 2026.
He is also a Founding Consortium Member of AIUC-1, an industry consortium defining how enterprises evaluate, insure, and underwrite AI systems. AIUC-1 is the world’s first AI agent standard covering data and privacy, security, safety, reliability, accountability, and societal impact, requiring independent third-party audits and quarterly adversarial testing across over 1,000 enterprise risk scenarios. Read Inside the OWASP AIVSS.
David contributed to the CISA Joint Cyber Defense Collaborative (JCDC.AI) initiative, participating in tabletop exercises that informed the AI Security Incident Collaboration Playbook released in January 2025. A second tabletop exercise was hosted at Scale AI’s San Francisco headquarters in September 2024 with approximately 150 participants from government and industry. Read CISA, JCDC, Government and Industry Partners Publish AI Cybersecurity Collaboration Playbook.
David is a coauthor of the WMDP Benchmark (Weapons of Mass Destruction Proxy), published in March 2024. The benchmark comprises 3,668 multiple-choice questions measuring hazardous knowledge in biosecurity, cybersecurity, and chemical security. Developed by a consortium of 56 academics and technical consultants including researchers from Scale AI and the Center for AI Safety, the benchmark was recognized by the White House Executive Order on Artificial Intelligence. Read The WMDP Benchmark: Measuring and Reducing Malicious Use With Unlearning.
Prior to Scale AI, David was a Staff Software Engineer at DoorDash between November 2019 and November 2022, focusing on cybersecurity and cloud platform engineering. Between April 2018 and October 2019, he was a Staff Security Software Engineer at Uber in San Francisco, where he had previously served as Technical Lead Manager from April 2015 to June 2017. He was a Senior Software Engineer at Google from January 2014 to April 2015 in Palo Alto, California, and concurrently served as Senior Software Engineer for Infrastructure Tooling at Nest Labs from November 2012 to April 2015. Read AI Security Isn’t Bullshit. But We’re Securing the Wrong Thing.
David began his career as a Service Engineer at Tellme Networks and Microsoft from March 2009 to May 2012. He earned his degree in Computer Science from Drake University in Des Moines, Iowa. He served as a Generative AI Industry Panelist at the Kelley School of Business at Indiana University and was an Advisor at Break Through Tech between May 2024 and April 2025, working to increase diversity in technology education.
Between August 2017 and August 2020, he volunteered as a Technical Startup Mentor at Fast Forward, working with five nonprofit startups at a mentorship event hosted by BlackRock. David is an internationally recognized keynote speaker on AI security and red teaming. Featured upcoming and recent speaking engagements include the AI Risk Summit 2025, NVIDIA GTC 2025, ODSC AI East 2026, DeveloperWeek, Devnexus, and the AI DevSummit.
His signature presentations include “AI Security Isn’t Broken. Our Mental Model Is” which reframes AI security around systems, identity, and trust for agentic AI, and “Domain-Limited General Intelligence: Before Things Go Too Far,” proposing safer evolutionary approaches between narrow AI and AGI. Watch AI Risk Summit Sessions.
David also operates as an independent AI security advisor, accepting two to three engagements per quarter. His advisory services include strategic AI security assessments for enterprises preparing to deploy AI systems, executive AI security workshops for C-suite briefings and team education, and ongoing strategic advisory engagements. His work has helped clients in financial services identify critical vulnerabilities pre-launch, achieve healthcare compliance for AI deployments, and establish AI security certifications for enterprise sales. David is based in the Boston, Massachusetts area.
Visit his LinkedIn profile, Personal Homepage, Medium page, and Sessionize profile. Follow him on Instagram, X, and YouTube.