Privacy, threat modelling, and secure AI delivery
View All TagsArtificial intelligence leadership is often misunderstood as the ability to select the best model, approve an AI strategy or sponsor a portfolio of proofs of concept. Those activities matter, but they are not the essence of leadership.
AI leadership is the disciplined conversion of uncertain technological capability into measurable, secure, governed and socially acceptable outcomes.
Deploying a large language model is not simply a question of choosing between an open-source model and a commercial API. It is an enterprise risk decision involving:
Before production, MonGo must show the hybrid customer-service AI is lawful, fair, secure, accountable, controllable and fit for purpose. The goal is not zero risk—it is proportionate controls, measured residual risk and evidence for every decision to continue, restrict or stop.
This article is Part VI of the Banking Customer-Service AI playbook. It follows Part I through Part V.
Most AI engagements fail not because the organisation lacks frameworks, but because it has too many of the wrong kind. Teams accumulate strategy canvases, maturity models, scoring formulas and governance checklists until the methodology itself becomes the delivery risk. The practical response is not a larger catalogue. It is a core set: enough structure to run end-to-end AI solution engineering, and little enough that practitioners can actually use it.
A Big Four firm does not compete only through its technical expertise, global network, technology platforms or client relationships. Its most valuable asset is trust.
Clients trust the firm with commercially sensitive information, strategic decisions, financial records, personal data and complex regulatory matters. Regulators trust the firm to exercise professional judgement and uphold required standards. Investors, employees, governments and the wider public expect the firm to behave responsibly, independently and ethically.
Data and AI create significant opportunities for growth, operational improvement, better decision-making and new client services. However, these opportunities also introduce risks that are different from, and often more complex than, those associated with traditional technology.
Information security for AI should not exist as a pile of threat models, vendor questionnaires and one-off pen tests. ISO/IEC 27001 establishes an organisation-wide Information Security Management System—scope, policy, risk assessment, control selection, operation, audit, management review and continual improvement—so confidentiality, integrity and availability remain enforceable when models, retrieval systems, agents and suppliers change.
The OWASP Top 10 for LLM Applications is not a reading list. It is a release language: every generative-AI system that touches tools, retrieval, personal data or production users should map risks, implement controls, test them and accept residual risk before scale.
A software licence is a legal agreement that gives a person or organisation permission to use software in ways that would otherwise be controlled by copyright law. The copyright owner is the licensor, while the person or organisation receiving the permissions is the licensee.
Artificial intelligence systems introduce a new class of security problems. A conventional application may fail because of insecure code, weak authentication or an exposed API. An AI application can suffer from all of those problems plus prompt injection, poisoned retrieval data, unsafe tool execution, model extraction, sensitive-data leakage, misleading outputs and autonomous agent behaviour.
AI red teaming is the structured practice of testing these systems from an adversarial perspective. The objective is not simply to make a model produce an inappropriate answer. It is to discover how an attacker, careless user, compromised data source or unexpected interaction could cause the complete AI system to violate its security, safety, privacy or business requirements.