PowerPoint Presentation

DIGITAL TRANSFORMATION LIFECYCLE Group Assignment Group 4.

SaaSpocalypse: Roughly$300 billion in combined market value lost across software-related indexes during the AI shock period (Wall Street Journal).

Contents. 01. Malikah Meyer. Scope of AI Impact. 02.

01. Define the Scope of AI Impact. Where AI can and cannot be applied in financial services.

DEFINING THE SCOPE OF AI IMPACT. AI is an enabler, not a standalone transformation strategy Delivers value when applied to well-defined, bounded problems Impact is strongest in: Data-intensive processes Pattern recognition and prediction Automation of repetitive activities Must operate within governance, regulatory, and ethical boundaries.

Operational efficiency Intelligent automation (RPA + AI) Incident prediction & anomaly detection Intelligent batch scheduling & workload optimisation Risk management and compliance Fraud detection and AML monitoring Credit risk modelling and transaction surveillance Customer & Adviser experience Chatbots and virtual assistants Personalised product recommendations.

WHERE AI IS CONSTRAINED (USE WITH GUARDRAILS). Credit Pricing and Underwriting Constraints: Explainability requirements, Regulatory scrutiny (e.g. fairness, transparency) Role of AI: Decision support, Scenario modelling, Risk segmentation Not acceptable: Black-box, fully automated approvals without explainability Financial Advise & Customer Outcomes Constraints: Act as an accountable decision maker, override suitability rules, provide unsupervised advice Role of AI: Support advisers with insights, summarize customer history, highlight risks and suitability gaps.

Fully Autonomous Decision-Making in High-Impact Areas Regulation of Automated Advice: FAIS Act & Fit and Proper Requirements (2017) mandate human oversight, algorithm monitoring, strong controls, governance, and adequate tech infrastructure; applies to robo-advisors. Consumer Protection & Principles: General Code of Conduct enforces disclosure, conflict of interest management, risk oversight, and technology-neutral standards. Future & Prudential Oversight: CoFI Bill to regulate digital innovation based on outcomes; Prudential Authority ensures AI risk management, operational resilience, and systemic risk mitigation..

Risk-based approach aligned to use-case materiality Mandatory controls: Data governance and privacy (POPIA) Model validation and monitoring Explainability (e.g. SHAP, LIME) Clear accountability: Human-in-the-loop decision models Board and executive oversight Transparency and disclosure for consumer-facing AI.

KEY TAKEOUTS. Checklist RTL Open folder Web design Decision chart Maze Processor 01 02 03 04 05 06 In operational efficiency, insight generation, and risk detection & prevention AI is most valuable Requires clear boundaries, not blind adoption Successful DT AI is not a business strategy in its own right AI ≠ strategy Where regulation, ethics, and accountability dominate AI is constrained Trust is the real transformation currency Financial Services Find relevant use cases that will deliver relevant business value and benefit Proper Scoping.

Image result for robot images hd png 02 Lifecycle Models(Waterfall, Agile, Spiral) vs AI.

The Core Shift. THE SHIFT We have moved from engineered precision to – rapid, intuition – based generation where prototypes are built almost instantly THE FAILURE MODE Traditional life cycle models rely on predictability that may no long exist. THE CONSEQUENCE Processes supporting AI must embrace ongoing evolution, accounting for variable outputs and environmental changes. However without a new model, speed becomes chaos. “I don’t write code anymore. I just let the agent write the code.” – Dario Armodei, Anthropic.

The Digital Transformation Spiral Model vs AI. THE LOGIC Separate the fast paced innovation from formal governance THE BENEFIT Teams (People and AI Agents) iteratively code on the Purple Loop, while the Green Loop ensures business value and safety Source: Adapted from Prof. Barry Myburth (DTSM).

EXPLORE Phase – Rapid Experimentation Focuses on fast, low-cost AI prototypes and quick feedback to enable high-velocity learning and insight generation EXPAND Phase – Scaling AI Teams (People and AI Agents) iteratively code on the Purple Loop, while the Green Loop ensures business value and safety EXTRACT Phase – Trusted Automation Optimizes AI systems for reliable, efficient output, building on trust and mature governance to enable stable automation..

Related image 03 Session Three. Definition-First Development.

AI accelerates outcomes — it rewards discipline..

Definition-First Development. AI is an amplifier of design discipline, not a substitute for it. Definition includes: Business outcomes Architecture context System context and patterns Test expectations Security/compliance Development Patterns Contract-First Development Blueprint before bricks. Guardrail-Driven Development Speed limits before acceleration. Test-Anchored Development Finish line before the race..

AI-Augmented, Definition-Driven. Understand → Design: AI is an amplifier not a substitute Workshops define: Domain boundaries Business capabilities Guardrails Risk tolerance Transform: AI becomes an operational capability Orchestration = coordinated change across: People, Process, Technology, Governance AI Orchestration: Embedded into workflows Integrated into lifecycle stages Measured and monitored Versioned and governed Govern →Learn →Iterate In the Spiral Model, review closes the loop. With AI, that loop must tighten, not loosen..

Challanges. 01 Team Balance is difficult Mismatched Velocity 02 Engagement / Belief It's all or nothing 03 Control 04 Adapt or Die.

04 Orchestration & Control. Related image.

AI Spew and Why Orchestration Matters. AI can produce high-volume and compelling outputs, but it may be incorrect. Newer chatbots answer everything, increasing probability of wrong answers that users cannot detect. Create operational noise and compliance risk Organisations need orchestration to keep AI fast and controlled.

Orchestration & Governance: The Control Tower. AI must operate as a governed system and not a standalone tool Orchestration ensures that AI operates using trusted, governed data, enforces policy rules, records all actions through audit logging, and routes outputs through the appropriate human approvals How does this tie back to the DTSM Outer Spiral: In the Transform phase, the organisation continuously monitors delivery progress, performance against Critical Success Factors, and emerging risks to ensure the AI initiative is delivering value while remaining controlled In the Evaluate phase, leadership decides next steps based on evidence, preventing hype-driven or unsafe scaling of AI.

Human in the Loop: Quality Control. Human in the Loop builds trust. Human oversight, judgment, and expertise are integrated into automated systems. AI outputs require adviser review and approval to ensure accuracy and ethical decision-making. This is the Quality Control handshake between AI and Human. Prevents harmful data outputs that could cause serious reputational damage and a loss of customer trust Quality Control may also detect: - Shifts in Prediction Drift - Hallucinations.

Minimum Governance Controls for AI. Four minimum Governance Controls: Continuous Monitoring, Transparency, and Audit Trails Risk‑based Human Approvals Data Quality, Lineage, and Privacy Controls for continuous improvement AI Inventory and Mapping Enables fast‑but‑safe scaling of AI.

05 Competitive & Extinction-Level Risk. Related image.

AI Adoption Is NOT Optional. AI fundamentally changes the cost and speed of financial services delivery Cost per transaction collapses and time to market compresses Personalization becomes the baseline, not a differentiator. Underwriting, pricing and servicing become real-time and data-driven Competitive advantage shifts to organization that redesign there operating model around AI..

The Real Risk is Organizational Inertia. The main threat is not technology – it is slow organizational change. Traditional governance and approval structures slow learning cycles. Competitors compound advantage through faster experimentation and feedback. AI requires iterative, spiral-based delivery rather than linear programmes. Delayed transformation becomes a strategic risk..

Extinction Level Risk Is Subtle. Competitive decline appears gradually, not suddenly. Margins tighten while cost and delivery friction increase. High value talent moves to AI-first organizations. AI-enabled firms learn faster and improve faster. In AI markets, long term survival depends on learning speed..

Conclusion | AI Requires a spiral, not a project.

06 Future Role of Planners. Related image. Course Area: Jobs-to-Be-Done / Organisational Change.

Role Evolution Framework. Future Role of Planners.

07 Synthesis & Conclusion. Related image. Course Area: Continuous Improvement / DTSM Loop.

Synthesis & Conclusion. Artificial Intelligence icon Artificial Intelligence icon Artificial Intelligence icon Acceleration Increases Governance Needs "Controlling highly autonomous systems" is key risk — Hassabis & Amodei Governance is essential risk management, not overhead AI Accelerates Transformation AGI possibly 2026-2027; 5x productivity gains already realized at SAP Annual planning cycles may be too slow Spiral-Based Continuous Change Required Cannot implement AI once and stop; each iteration expands capability and revisits risk Alam's stages Empower →Confidence →Autonomy are spiral cycles.

Thank You.

ACADEMIC REFERENCES. Financial Sector Conduct Authority (FSCA) & Prudential Authority (PA). (2025). Artificial Intelligence in the South African Financial Sector. South Africa. Organisation for Economic Co-operation and Development (OECD). (2024). Regulatory Approaches to Artificial Intelligence in Finance. Bank for International Settlements (BIS). (2024). Regulating AI in the Financial Sector: Recent Developments and Main Challenges. World Economic Forum. (2025). Artificial Intelligence in Financial Services. Protection of Personal Information Act 4 of 2013 (POPIA), South Africa..