Staff Software Engineer - Agentic Ai Security

Commonwealth Bank Of Australia; Western Australia; On-site; AUD 140,000 - 220,000; Full time

Job summary

Commonwealth Bank of Australia seeks a senior, hands-on Identity Services engineer to lead the secure, scalable delivery of agent-based identity services. You will drive architecture decisions, mentor engineers, and shape the roadmap for agentic AI security across the enterprise.

You will collaborate with cross-functional teams to implement Entra ID for Agents, enforce zero-trust controls, and develop robust patterns for identity lifecycle and governance of non-human identities.

Qualifications

• Hands-on expertise with Microsoft Entra ID for Agents (Entra Agent ID) and Microsoft Agent 365, or comparable agent identity and governance platforms.
• Strong knowledge of Microsoft Entra ID, OAuth 2.0, OpenID Connect, application identities, service principals, managed identities and workload identities.
• Experience designing enterprise identity lifecycle and governance patterns for non-human identities.
• Applied understanding of agentic AI threats, threat modelling, trust boundaries, sandboxing, tool allow-listing, permission scoping and secure handling of untrusted content.
• Understanding of secure delegation, token exchange and emerging agent identity approaches such as AI Identity Management System (AIMS).
• Familiarity with Microsoft 365 Copilot, Copilot Studio agents, Microsoft Graph permissions, connectors or MCP-style tool integrations.

Responsibilities

• Lead engineering adoption of Microsoft Entra ID for Agents (Entra Agent ID) and Microsoft Agent 365, including agent registration, ownership, permissions, inventory, visibility and lifecycle controls.
• Establish secure engineering patterns for agent identity, authentication, authorisation, secrets, tool access and privileged operations.
• Design controls for prompt injection, excessive agency, unsafe tool use, credential exposure, privilege escalation and data leakage.
• Apply Zero Trust, least privilege, identity isolation, scoped permissions, short-lived credentials and human approval to agent architectures.
• Build and prototype secure agentic AI capabilities that improve Identity Services engineering and operational workflows.
• Ensure agent activity is observable, attributable, auditable and appropriately governed.
• Provide hands-on technical leadership, mentoring engineers and simplifying or retiring obsolete architecture.
• Shape the technical roadmap for agentic AI, automation and identity engineering within Identity Services.

Skills

Identity engineering

Security architecture

AI security

Leadership

Cloud security

Hands-on coding

Education

Bachelor's degree in Computer Science or related field

Tools

Microsoft Entra ID

Entra Agent ID

Microsoft Agent 365

OAuth 2.0

OpenID Connect

APIs

PowerShell

Python

Job description

• Shape how AI agents are securely identified, authorised and governed across a large-scale enterprise identity ecosystem.
• Take on a hands-on technical leadership role combining agentic AI security, identity engineering and emerging technology.
• Work at the forefront of investment in AI, automation and identity security to help build simpler, safer and more scalable engineering capabilities.

Do Work That Matters

AI agents create new possibilities for how we build, automate and operate technology, but they also introduce new identity and security challenges.

See yourself in our team

You’ll join our Identity Services engineering community, taking a senior hands-on technical role across a squad delivering secure, reliable and scalable identity services.

You’ll combine deep technical problem-solving with architecture and engineering leadership, helping engineers tackle complex challenges while influencing the longer-term direction of agent identity and agentic AI security.

Your responsibilities will include

• Lead engineering adoption of Microsoft Entra ID for Agents (Entra Agent ID) and Microsoft Agent 365, including agent registration, ownership, permissions, inventory, visibility and lifecycle controls.
• Establish secure engineering patterns for agent identity, authentication, authorisation, secrets, tool access and privileged operations.
• Design controls for prompt injection, excessive agency, unsafe tool use, credential exposure, privilege escalation and data leakage.
• Apply Zero Trust, least privilege, identity isolation, scoped permissions, short-lived credentials and human approval to agent architectures.
• Build and prototype secure agentic AI capabilities that improve Identity Services engineering and operational workflows.
• Ensure agent activity is observable, attributable, auditable and appropriately governed.
• Provide hands-on technical leadership, mentoring engineers and simplifying or retiring obsolete architecture.
• Shape the technical roadmap for agentic AI, automation and identity engineering within Identity Services.

We’re interested in hearing from people who:

Bring strong identity engineering expertise together with practical knowledge of AI agents and the security risks created when autonomous systems receive credentials, permissions and access to enterprise services.

You’ll be comfortable moving between architecture, code, APIs and identity platforms, and translating emerging security concepts into engineering controls that work in practice.

Ideally, you'll bring:

• Hands-on expertise with Microsoft Entra ID for Agents (Entra Agent ID) and Microsoft Agent 365, or comparable agent identity and governance platforms.
• Strong knowledge of Microsoft Entra ID, OAuth 2.0, OpenID Connect, application identities, service principals, managed identities and workload identities.
• Experience designing enterprise identity lifecycle and governance patterns for non-human identities.
• Applied understanding of agentic AI threats, threat modelling, trust boundaries, sandboxing, tool allow-listing, permission scoping and secure handling of untrusted content.
• Understanding of secure delegation, token exchange and emerging agent identity approaches such as AI Identity Management System (AIMS).
• Familiarity with Microsoft 365 Copilot, Copilot Studio agents, Microsoft Graph permissions, connectors or MCP-style tool integrations.
• Practical PowerShell, Python or similar coding skills, with confidence working across APIs, SDKs and configuration-as-code.
• The technical leadership skills to influence engineering direction, mentor others and take ownership of complex strategic initiatives.

Working with us

At CommBank, we're committed to creating an accessible, inclusive and respectful workplace. If you require support or adjustments, please let us know. We welcome applications from people of all backgrounds and we're particularly committed to making a positive difference for Aboriginal and/or Torres Strait Islander Peoples. For support please contact 1800 989 696.

We're aware of some accessibility issues on this site, particularly for screen reader users. We want to make finding your dream job as easy as possible, so if you require additional support please contact HR Direct on 1800 989 696.