Wednesday, November 18, 2026

4:00 PM - 5:00 PM

Bradman Theatre

---

Sponsored by

---

Presentation Outline

Artificial intelligence promises faster decision cycles, greater workforce capacity and new forms of operational advantage. In Defence, however, trust cannot be determined by model performance alone. AI must operate across complex estates, fragmented data, multiple security domains, hybrid environments and established chains of accountability. The central challenge is not simply adopting AI, it is providing AI with trusted context and tightly governed pathways to act. Salesforce’s presentation examines the enterprise foundations required to operationalise AI securely, ethically and at mission-speed and scale. It explores how governed integration and reusable APIs connect legacy, cloud, data and partner environments; how data quality, lineage, metadata and master-data management establish confidence in the information supplied to AI; and how identity, access controls, observability and human-approval boundaries constrain what AI agents can know and do. The session will show how our Data Foundations combine to create trusted organisational context, while Agentforce provides bounded reasoning and action within defined guardrails. It will also examine how #Missionforce applies these capabilities to national-security missions, including AI readiness, sustainment, workforce, logistics and delivery. Using a Defence-relevant readiness scenario, the presentation will trace the journey from fragmented operational signals to an evidence-based recommendation, authorised human decision and auditable action. It will address sovereignty, zero-trust access, resilience, lifecycle governance and operation across distributed environments without proposing unconstrained autonomy. Attendees will leave with a practical, architecture-led framework for adopting AI progressively: connecting approved systems, establishing authoritative data, defining machine authority, preserving human accountability and expanding capability only when operational evidence supports it.

---

Speaker