AI Security Architect

Description
About Our Team
The Product Security organization is seeking a dedicated and experienced AI Security Architect to design and advise on security strategies to protect our AI systems, data, and products. You will define the strategy and technical implementation of an AI/ML secure development lifecycle, integrated with our existing software development lifecycle

This is a dedicated security role that demands a unique blend of deep AI system understanding and a strong adversarial security mindset. The AI Security Architect will champion the integration of security-by-design principles into all AI initiatives, proactively identifying and mitigating AI-specific vulnerabilities to safeguard Salesforce's intellectual property, customer data, and brand reputation.

This crucial role will drive a unified and cohesive AI security strategy throughout the entire Salesforce ecosystem, transcending traditional product and "cloud" boundaries. Recognizing that agentic AI necessitates a fundamental shift from siloed approaches, the AI Security Architect will foster a consistent and integrated security posture across all Salesforce products, ensuring a shared commitment to customer trust by influencing platform design with critical security context.

Impact – Responsibilities

• Advise on Secure AI Platform Design & Guidance
• Define Unified AI Security Standards & Best Practices
• Work with BISOs to Drive Secure-by-Default and Trust Principles Across Clouds
• Provide Deep Technical Input into Technology Security Evaluation & Risk Assessment
• Ensure Cross-Alignment with Ethical AI, Legal, and Compliance
• Manage Cutting-Edge Technology Security

Advise on Secure AI Platform Design & Guidance

• Partner with engineering and development architects to advise on the secure architecture design for end-to-end AI systems, including data pipelines, model development and deployment (MLOps), and integration with existing enterprise systems. Provide security context, risk identification, and recommendations for scalable, reliable, and secure designs across all Salesforce clouds.

Define Unified AI Security Standards & Best Practices

• Establish and enforce security architectural standards, design patterns, and best practices for AI development. Focus on ensuring robust, secure, and explainable AI systems with a consistent, unified security approach across all Salesforce products and clouds.

Work With BISOs to Drive Secure-by-Default and Trust Principles Across Clouds

• Ensure that security is fundamentally ingrained in the design of all AI components from inception, reinforcing Salesforce's "Secure by Default" architecture and unwavering commitment to customer trust across the entire platform. This role will influence development practices to embed these principles.

Provide Deep Technical Input into Technology Security Evaluation & Risk Assessment

• Evaluate and recommend appropriate AI technologies, frameworks (e.g., TensorFlow, PyTorch), platforms (e.g., cloud AI services), and tools from a security perspective, ensuring they align with our security requirements and risk appetite for a multi-cloud enterprise.

Ensure Cross-Alignment with Ethical AI, Legal, and Compliance

• Work closely with legal, compliance, AI Research, and the Office of Ethical and Humane Use teams to embed responsible AI security practices consistently, and to align on consistent taxonomies, definitions, methodologies, and tooling for testing security, safety, and ethical alignment of AI models and systems. Champion ethical AI principles from a security standpoint, ensuring compliance with relevant data privacy regulations (e.g., GDPR, CCPA) and internal ethical guidelines across the entire Salesforce ecosystem.

Manage Cutting-Edge Technology Security

• Stay continuously informed with the rapid pace of change in generative AI and related emerging technologies. Work with engineering architects to strategize for securely operationalizing, developing, and integrating these new technologies into the platform to maintain Salesforce's competitive advantage in the marketplace. Drive innovation in security controls tailored to these evolving advancements.

Stakeholder Engagement & Ecosystem Alignment

• Act as a subject matter expert and trusted advisor on AI security architecture to senior leadership, engineering teams, product managers, and other stakeholders across all Salesforce products and "clouds," fostering cross-cloud alignment and a unified approach to AI security.

Work Towards Breaking Down Silos & Encouraging Collaboration

• Proactively break down organizational silos and facilitate seamless collaboration between different product teams and clouds regarding AI security, driving a singular, consistent AI security posture across the entire Salesforce platform by influencing development and deployment practices.

Fostering Security Integration & Cross-Functional Partnership

• Collaborate with the broader security organization (e.g., AppSec, Cloud Security, Incident Response) to integrate AI security into the overall enterprise security strategy and incident response plans, ensuring that cross-cloud AI security concerns are addressed comprehensively. This also includes close partnership with AI Research and the Office of Ethical and Humane Use to design security-related models, ensure comprehensive testing, and align on shared frameworks for ethical, safety, and security evaluation.

Active External Engagement

• Represent Salesforce at industry conferences and forums on AI security, contributing to the broader AI security community and showcasing Salesforce's leadership in secure and ethical AI deployment across diverse product offerings.

Risk Identification & Management of AI Products

• Lead comprehensive risk assessments across architecture, design, deployment, and runtime phases.
• Oversee technical reviews, threat modeling, code/design reviews, and hands-on testing to uncover and mitigate risks.
• Analyze diverse risk signals and discovery data to prioritize security activities and inform the product security roadmap.
• Evaluating trade-offs of risk remediation strategies, recommending optimal solutions that balance security, functionality, and business objectives.
• Evaluate and integrate responsible AI principles—including data governance, model integrity, explainability, and adversarial resilience—into core Agentforce engineering practices.
• Collaborate with Security Customer Enablement, Privacy, and Legal teams to define secure AI usage frameworks and external trust communications.
• Guide engineering teams through architecture reviews, code analysis, and red-team feedback loops for model-driven products.

Minimum Qualifications

• A related technical degree required.
• First-hand work with Machine Learning, Deep-Learning, or Artificial Intelligence.
• Familiarity with current attacks on ML models, including adversarial examples, training data extraction, model extraction, and data poisoning.
• Proven experience in developing new attacks and defenses for ML/AI enabled applications.
• Proven communication, collaboration, and interpersonal skills with the ability to effectively communicate complex technical concepts to diverse audiences, including technical and non-technical teams.
• An attacker's mindset; consider abuse and attack paths as well as the defensive mindset to recommendations to prevent them
• A passion around improving the security development lifecycle and delivering security guidance to engineers in a language they understand.
• Ability to work with data, identify trends and propose comprehensive mitigations that eradicate systemic security concerns.
• Experience participating in an information security program and improving or proposing improvements to a secure development lifecycle
• Threat modeling of security topics across infrastructure security & application security domains Understanding of TCP/IP, common networking ports and protocols, traffic flow, system administration, OSI model, defense-in-depth and common security elements
• Excellent writing and presentation skills.
• Possess the ability to communicate concisely, clearly, and intelligently to cross functional teams.

Preferred Qualifications

• Proven ability to drive enterprise-ready features to release prioritizing Security and Availability without sacrificing usability
• Strong technical aptitude; comfortable engaging with engineering on architecture, APIs, and platform implications
• Experience working with InfoSec, Legal, and IT stakeholders at Fortune 500 companies
• Demonstrated expertise in navigating security incident response efforts, including direct engagement with executive leadership, legal counsel, and external parties to resolve complex security issues.
• Strategic thinker with an understanding of the evolving global threat landscape and its implications for millions of users.

Additional Exceptional Qualifications

• Experience with client side/browser security features like same origin policy, CORS, CSP, shadow DOM, Web Components, web development frameworks etc.
• Experience with software development in one or more languages such as: JavaScript, Java, Python, Ruby, PHP, Go, TypeScriptSome experience performing penetration testing or familiarity with the process
• 5+ years proven experience in the following areas in a security engineering or research role:Securing products and infrastructure from the OWASP Top 10 and/or CWE Top 25
• Exploiting web and web services security vulnerabilities such as cross-site scripting, cross site request forgery, SQL injection, DoS attacks, XML/SOAP, API attacks, etc.Public Cloud security architecture in one or more of the following: Amazon Web Services, Google Cloud Platform, Microsoft Azure, Alibaba Cloud, etc.

For roles in San Francisco and Los Angeles: Pursuant to the San Francisco Fair Chance Ordinance and the Los Angeles Fair Chance Initiative for Hiring, Salesforce will consider for employment qualified applicants with arrest and conviction records.