Press ALT + F8 to see a list of keyboard shortcuts

Discussion board 11.3: Think & share šŸ’”Leading Responsible Agentic AI Transformation in HR and Finance

All Sections
4 unread replies. 4 replies.
  1.  Module 11: Introduction and Instructions (Completed)
  2.  Video 1 [02:49]: Module Introduction (Completed)
  3.  Video 2 [06:49]: Why Agentic AI Matters in HR & Finance (Completed)
  4.  Video 3 [06:52]: Agentic AI in Hiring & HR Operations (Case Studies I) (Completed)
  5.  Self-Study Quiz 11.1: Think & ApplyšŸ’”Recognising High-Impact Opportunities for Agentic AI in HR & Finance ((Must submit the Quiz))
  6.  Video 4 [05:01]: Scaling Hiring Intelligence with Agentic Systems (Case Studies II) (Completed)
  7.  Video 5 [05:02]: Agentic AI in Finance: From Analysis to Augmentation (Completed)
  8.  Video 6 [05:01]: Real-Time Finance: Fraud Detection and Strategic Shift (Completed)
  9.  Video 7 [06:01]: Accountability in Agentic Systems (Completed)
  10.  Self-Study Quiz 11.2: Think & ApplyšŸ’”Evaluating Agentic AI Adoption, Governance and Strategic Readiness ((Must submit the Quiz))
  11.  Video 8 [05:07]: Strategic Implications: Rethinking Operating Models (Completed)
  12.  Video 9 [05:44]: From Deployment to Value: Implementation, Risks and Metrics (Completed)
  13.  Discussion board 11.3: Think & share šŸ’”Leading Responsible Agentic AI Transformation in HR and Finance ((Not yet completed))
  14.  Video 10 [03:54]: Final Thoughts ((Must view in order to complete this module item))
  15.  Self-Study Assignment 11.4: Think & ActšŸ’”Developing an Executive Roadmap for Agentic AI Adoption ((Must view in order to complete this module item))
  16.  Reading: Executive Readiness Check: Scaling Agentic AI ((Must view in order to complete this module item))
  17.  Self-Study Assignment 11.5: Think & ActšŸ’”Evaluating Organisational Readiness for Responsible Agentic AI Adoption ((Must view in order to complete this module item))
  18.  Summary and Video Transcripts: Module 11⭐ ((Must view in order to complete this module item))
  19.  Professor Ian's Recommendation for Further Learning: Module 11 ((Must view in order to complete this module item))
  20.  Module 11: Q&A Discussion Board ((Not yet completed))
  21.  Module 11: Feedback Survey ((Must submit the Quiz))

Module Outcomes.pngLearning Outcomes Addressed: 

  • Evaluate accountability frameworks that reduce risk and prevent failure in AI-led initiatives.
  • Apply a practical management checklist to guide adoption and scaling within HR and finance teams.

Scenario

Your organisation is preparing to expand Agentic AI across several HR and Finance functions following a successful pilot. While leadership is enthusiastic about the potential benefits, there are differing views on how quickly AI should be scaled and the level of human oversight that should be maintained.

As a senior leader, your role is to balance innovation with accountability, ensuring AI delivers measurable business value while maintaining trust, transparency and responsible governance.

Reflecting on the accountability principles, strategic implications and implementation guidance presented in this module, discuss the following:

What leadership decisions are most critical to ensuring Agentic AI delivers long-term organisational value without compromising accountability, employee trust or business outcomes?

Support your response with examples from your own organisation or a workplace you are familiar with.

In your discussion, consider:

  • Why AI adoption should be viewed as an operating model transformation rather than simply a technology implementation.
  • How leaders should determine which decisions require human oversight.
  • The organisational capabilities that should be established before scaling Agentic AI.
  • The risks of expanding AI too quickly without demonstrating measurable business value.

We also encourage you to go through your peers' responses and share your views.

Estimated Duration: 30 minutes

NOTE: This is a practice activity and will not be graded or impact your programme completion. However, we strongly encourage participation for a more holistic learning experience.

Filter replies by unread    
Reply to Discussion board 11.3: Think & share šŸ’”Leading Responsible Agentic AI Transformation in HR and Finance
Cancel
  • Collapse Subdiscussion Priya Jha
    • Why AI adoption should be viewed as an operating model transformation rather than simply a technology implementation.

    Technology is only a part of any transformation program, same goes for AI adoption. We need to conisder how agents would interact with existing teams, processes, data and decision making structures.  I would go for redesigning the operating model around responsible automation, rather than simply slopping AI over existing processes.

    • How leaders should determine which decisions require human oversight.

    I would consider three factors: Impact , Reversibilit & Confidence and evidence

    Human oversight has to be risk-based, not simply applied equally to every AI workflow.

    • The organisational capabilities that should be established before scaling Agentic AI.

    I believe organisations should establish several foundational capabilities like Clear accountability , Data governance, Human oversight, Monitoring and auditability, Employee enablement & Continuous evaluation

    • The risks of expanding AI too quickly without demonstrating measurable business value.

    For Agentic AI, I would start with clearly defined use cases, measure results, learn from failures and scale based on evidence rather than enthusiasm.

    Reply to Comment
    Cancel
    • Collapse Subdiscussion Edoardo Bertolani

      I see Agentic AI transformation primarily as an operating-model redesign, not simply a technology implementation. This is particularly relevant in HR and Finance because both functions ultimately manage resources — human and financial — where efficiency, consistency and optimisation are generally more important than creativity.

      For this reason, I am not convinced that permanent human approval of consequential decisions should be the ultimate model. If two or three genuinely independent and capable AI systems analyse the same problem, challenge the information and assumptions, and independently reach the same conclusion, why should the judgement of a single human supervisor automatically be considered more reliable?

      I see the future model moving from human-in-the-loop to human-on-exception. Independent AI agents could perform and cross-check the analysis, while humans intervene when agents disagree, information is incomplete, confidence is insufficient, or exceptional circumstances are identified.

      From my banking experience, I would therefore scale gradually, but with the objective of increasing autonomy as performance is demonstrated. The essential foundations are reliable data, independent AI validation, auditability, measurable quality thresholds and clear escalation rules. The real question should eventually become not ā€œWhere must a human approve the AI?ā€ but ā€œUnder what circumstances does the evidence show that human intervention actually improves the decision?ā€

      Reply to Comment
      Cancel
      • Collapse Subdiscussion Nora Pataut

        Key Considerations for Scaling Agentic AI

        • Operating Model Transformation vs. IT Upgrade: Treating Agentic AI merely as software installation leads to siloed adoption. Leaders must view it as an operating model transformation that fundamentally redesigns HR and Finance workflows, redefining how human teams and autonomous agents collaborate, make decisions, and share accountability.

        • Defining Human-in-the-Loop Thresholds: Establish a clear governance matrix to determine where human oversight is mandatory. Routine actions (e.g., automated expense categorization) can operate autonomously with periodic auditing, while high-impact, value-sensitive decisions (e.g., compensation adjustments, compliance reporting, or hiring recommendations) require strict human-in-the-loop sign-offs to preserve trust and compliance.

        • Foundational Organizational Capabilities: Before scaling, organizations must build critical capabilities, including robust data governance, cross-functional training (upskilling employees to manage and audit AI agents), and unified risk/compliance frameworks. Without these guardrails, scaling accelerates systemic errors and vulnerabilities.

        • The Dangers of Premature Scaling Without Measurable Value: Expanding AI too rapidly without establishing clear baseline metrics risks creating hidden technical debt, employee resistance, and unjustified operational costs. Proof of concept metrics (e.g., cycle-time reduction, accuracy rates) must be transparently demonstrated before securing broad budget allocation for enterprise-wide deployment.

        • Balancing Innovation Velocity with Employee Trust: Rapid automation can spark employee anxiety around job security and transparency. Leadership must prioritize clear communication regarding how AI augments rather than replaces human work, ensuring psychological safety and sustaining organizational trust throughout the transition.

        Reply to Comment
        Cancel
        • Collapse Subdiscussion Hadrian

          I look at this triangle of business operation - People, Process, Performance - holistically, when implementing AI in HR and/or Finance function

          People is critical part on AI implementation in longer term, i.e. 

          • Does people within the functions have capability on AI now, or plan is in place to up-skill people capability, or plan to have the right talent. All should lead into overall capability of people to work in collaboration with AI
          • Which part of HR or Finance will be impacted the most on number of people as AI implemented - at the same time do we have the right design of organization to continue operate, to ensure sub function that handle most consequences not weaken. 
          • As HR and Finance implement AI earlier than other function, have we define - 5 steps in advance - the future skills that are critical for these 2 functions, as both will keep progressing ahead of other function. 

          Process should be prioritized, AI implementation must not exposed high risk process, where human intervention is needed. On the other hand, a success in least risk process will give confidence to implementation in more process. Within HR and Finance, while operations are mainly process based and repetitive, in most cases inter-connected with other functions or some process can be very sensitive. It is impacting the clear governance into the process as well. Some example below:

          • Employee performance review and rating are annual process that are handled by HR, while in most cases the process are clear - employee better achievement on target = better rating. But there are some qualitative judgement need to be attached, including 360 degree view from other colleagues. it is definetely need human judgment within the process.
          • Finance team lead the S&OP process together with sales team, marketing team and other commercial team. Any decision on the demand, sales target can be easily model by AI and to be reviewed cross function every month. However, final decision need more human judgement, in many occasion can't be factored into the model

          Performance of the business will be the ultimate decision if the AI implementation will bring positive impact to the business. It will move the end game from OUPUT KPI of AI implementation, like efficiency, adoption rate, quality strengthening, to be OUTCOME KPI - what does the impact to business growth?

          Reply to Comment
          Cancel