Security Architect at a Major Japanese Bank
Description Al Security Architect Remote Opportunity The Al Security Architect will support the Al Security transformation program The consultant will provide hands-on leadershi technology teams, vendors, and regional security leaders. The position is accountable for moving work from analysis through decision, implemen improper output handling, excessive agent autonomy, model and Al supply chain risk, vector / embedding and retrieval-related weaknesses. Part for model traffic, including prompt inspection, response filtering, PII/secret redaction, model access control, rate limiting / abuse prevention, audit Comments The Al Security Architect will support the Al Security transformation program The consultant will provide hands- on leadership across Al security architecture, prompt and data protection, shadow Al discovery, model security, monitoring, governance, and platform reuse. The role will work across AD, EMEA, APAC, and HO with security architecture, engineering, operations, finance, procurement, risk, compliance, legal, privacy, business technology teams, vendors, and regional security leaders. The position is accountable for moving work from analysis through decision, implementation, evidence, and sustainable operational adoption. Skills Lead security initiatives for LLM-powered applications, copilots, agentic systems, and Al-assisted workflows.This is an hands on role. Design and implement controls to reduce the risks of prompt injection, sensitive information disclosure / data leakage, improper output handling, excessive agent autonomy, model and Al supply chain risk, vector / embedding and retrieval-related weaknesses. Partner with engineering teams to embed secure-by-design patterns into Al application development, deployment, and operations. Create guardrails for Al systems that process customers, employees, and regulate financial data. Architect and implement Al gateway controls that centralize security policy enforcement for model traffic, including prompt inspection, response filtering, PII/secret redaction, model access control, rate limiting / abuse prevention, audit logging and evidence generation. Proficiency in at least one modern scripting or programming language (e.g., Python, Go, or Node.js) for automation, tooling, and integrations.