DevoteamInsightsExpert ViewWill Product Owners and Scrum Masters Survive the AI Boom?

Will Product Owners and Scrum Masters Survive the AI Boom?

Authors

  • Gianluca Odoardi
    Senior Project Manager – Career Manager – AI Tech Lead

  • Diego Valentierra
    Product Owner, Coach Agile/Scrum Master

Share

“It is not AI that redefines our work, but the way we choose to integrate it”

For over thirty years, Agility has established itself as the most relevant answer to the challenges of market complexity and speed. Born from the desire to break free from overly rigid waterfall methods, it has profoundly transformed how organisations design, develop, and deliver products.

Rooted in collaboration, iteration, and adaptation, Agility has established a true culture of flexibility and continuous value. It is based on now well-established principles, those of the Agile Manifesto, which value individuals and their interactions more than processes, customer collaboration over contract negotiation, and responding to change over following a fixed plan. Structuring approaches, such as the Scrum framework, the Kanban methodology, or Agile models like SAFe, combined with continuous improvement cultures like Lean and DevOps, have enabled the formalisation and dissemination of these practices across multiple sectors, including banking, healthcare, and the pharmaceutical industry. The success of Agility lies in its ability to make teams more autonomous, transparent, and reactive, with measurable gains.

However, today, a new force is shaking these foundations: Artificial Intelligence (AI) and its capacity to disrupt the very foundations of Agility. As a catalyst for transformation, it introduces task automation, content generation, and a profound change in decision-making. This emergence raises fundamental questions for Agility and its key players, particularly Product Owners, Scrum Masters, and developers.

Indeed, industry forecasts confirm this transformation. Gartner predicts that by 2030, 80% of transactional project management tasks (such as tracking and reporting) will be executed by AI. Concurrently, a Project Management Institute (PMI) survey revealed that the share of projects managed by AI is expected to increase from 23% to 37% in just three years (Isakova, 2021). Beyond the simple automation of repetitive tasks, AI is establishing itself as a strategic and cognitive player within teams. It generates content (code, documentation, and user stories), analyses data to inform decision-making, and suggests backlog priorities based on predictive models.

This technological irruption raises fundamental questions for Agility and its pivotal roles, particularly Product Owners (PO) and Scrum Masters (SM). If AI can automate a significant portion of backlog management, specification writing, or team metric analysis, the central question remains. What is the unique and irreplaceable value proposition of these roles in the future? If their relevance no longer lies in task execution, it will gain value in the ability to reinvent themselves, that is, to master and guide the ethical and strategic integration of this technology in the service of value creation and collective performance.

The purpose of this article is to explore how generative AI challenges the pillars of Agility and, by extension, redefines the role of the central figures of these methodologies, pushing them to reinvent themselves to ensure their development and evolution in this new era of innovation.

In this article, you’ll read:

The Evolution of the PO and SM roles since the Publication of the Scrum Guide

The roles of Product Owner and Scrum Master, although initially defined in the Scrum Guide, have become roles commonly adopted by many teams engaged in an Agile approach (even if these roles are not required by approaches like Kanban). However, they have evolved over time to meet market realities and new technological challenges.

The Product Owner: From Execution to Strategy

Historically, the Product Owner’s role focused on clear and operational responsibilities. The PO was the sole and primary representative of the customer’s voice and stakeholders within the development team. Their missions traditionally centred on:

  • Managing and prioritising the Product Backlog. The PO is the guardian of the Backlog. They ensured that epics, features, and user Stories were well-defined, clear, concise, and, above all, ranked by business value and risk for the company. This prioritisation was an ongoing exercise of balancing user needs, technical constraints, and the organisation’s strategic objectives.
  • Defining the product vision and strategy. In close collaboration with Product Management, the PO helped translate the overall strategy into an actionable roadmap. They communicated this vision in an inspiring way to the development team to maintain alignment and motivation.
  • Accepting deliverables. At the end of each sprint, the PO is the only one who can validate and accept that the developed increments met the Definition of Done. And, more importantly, the initially specified business acceptance criteria. They are therefore responsible for the functional quality of the delivered product.
  • Accepting deliverables. At the end of each Sprint, the development team was responsible for ensuring that the developed increments met the Definition of Done. Thereby ensuring technical quality (tests, code reviews, integration, etc.). The Product Owner, for their part, was the only one who could accept the increment, validating that it met the initially specified business acceptance criteria. They were therefore responsible for the functional quality and business value of the delivered product.
  • Collaboration and communication. They acted as the bridge between the business and the domain. This involved regular backlog refinement meetings with the team and demonstration sessions with stakeholders.

These responsibilities defined the PO as a key functional decision-maker. Success relied heavily on their communication and negotiation skills, as well as their deep knowledge of the business domain and users. With the updates to the Scrum Guide, particularly the 2020 version, the role has gained depth. The introduction of the Product Goal concept has prompted the PO to commit to a long-term vision for the product, extending beyond the backlog. It is no longer just about knowing what to do, but about defining why we are doing it. This has increasingly positioned them as a product leader and a strategic decision-maker.

The Scrum Master: From “Servant Leader” to Change Coach

The Scrum Master role, as initially defined in the Scrum framework, is fundamentally based on the concept of a “servant leader,” which corresponds to a leader who serves the team. This designation is not merely a title, but an essential posture that dictates all their actions and interactions. As a servant leader, the Scrum Master has the primary responsibility of serving the development team, the Product Owner, and the organisation as a whole. This involves:

  • Facilitate and guide. He does not give orders, but facilitates the Scrum events (Sprint Planning, Daily Scrum, Sprint Review, Retrospective). He guides the team in the correct application of the values and principles of Agility and Scrum.
  • Remove impediments. His primary mission is to identify and remove all obstacles (technical, organisational, and human) that slow down or prevent the team from achieving its Sprint goals.
  • Coach the team. He is a coach for the team, helping it to self-organise, become cross-functional, and continuously improve (via Retrospectives).
  • Protect the team. He acts as a buffer between the team and non-productive external interferences. Thus, ensuring that the team can fully concentrate on value creation.
  • Serve the organisation. He works to spread the Agile culture within the organisation. He helps to transform the working environment to make it more conducive to Agility.

His legitimacy, therefore, does not come from hierarchical authority but from his ability to create the optimal conditions for the team to deliver value effectively and sustainably. He is the guarantor of the Scrum framework. Over the years, the SM’s role has moved beyond simple facilitation. He has become a true team coach, responsible for the effectiveness and self-organisation of the Scrum Team. The 2020 Scrum Guide even removed the prescriptive questions from the Daily Scrum, reinforcing the idea that he must coach the team to find its own solutions, rather than just facilitating a meeting. This professional has also extended his scope of action to the organisation as a whole, acting as a change agent to help the company adopt Agile principles.

The challenge that AI Poses to these Two Roles in Daily Product Development

Generative artificial intelligence is no longer content to be a simple decision-making aid. It is now intruding into the daily life of product development. AI acts increasingly as a digital “teammate,” capable of automating tasks, generating insights, and disrupting collaboration. These evolutions have concrete impacts on both roles, requiring a re-evaluation of their actions, priorities, and required skills.

For the Product Owner: AI as a Co-creator of Value

The Product Owner must now partner with a capable companion that can analyse enormous volumes of data faster than any human. Indeed, AI can:

Generate user stories and technical specifications

From a simple description of a need, AI can draft complete user stories, including detailed acceptance criteria. The PO no longer spends their time formulating every sentence, but rather validating and refining the AI’s suggestions. This vigilance is essential given the risk of hallucinations. A generated user story may indeed seem coherent on the surface, while containing subtle inconsistencies or unrealistic acceptance criteria. This risk only reinforces the Product Owner’s role in validation. Despite this necessary supervision, the time savings on drafting remain significant. An academic study conducted by Santos and his collaborators in 2025 showed an efficiency rate of 88.57% for user story generation via advanced prompting techniques.

Analyse the market and customer feedback

AI tools can scan thousands of customer feedback reports, online reviews, or market studies to extract trends and feature opportunities. The PO must transition from collecting information to strategically interpreting this data. They must focus on the “why” and not just the “what.”

Analyse the market and customer feedback

AI tools can scan thousands of customer feedback reports, online reviews, or market studies to extract trends and feature opportunities. The PO is responsible for both aspects:

  • They define the “why” (business value)
  • They are the final decision-maker for the “what” (Backlog content)

AI primarily helps them generate the potential “what”. This allows them to spend more time on the strategic interpretation of this data and on the strategic “why”. But they do not, in any case, relinquish final responsibility for the Backlog.

Prioritise the backlog with predictive data

By analysing product usage data, AI can suggest features that will have the greatest impact on engagement or conversion. The PO is then challenged by data-based proposals, forcing them to justify their choices not only with their intuition but also with an in-depth analysis of metrics. Furthermore, quantitative benchmarks support this direction. An empirical study by Oftebro and colleagues in 2025 demonstrated a reduction of approximately 45% in completion time for backlog refinement tasks thanks to GenAI.

The emergence of autonomous agents

Beyond simple content generation, the arrival of autonomous agents like AutoGPT or Devin enables the autonomous pursuit of complex objectives, such as decomposing an Epic into technical tasks. In this context, the Product Owner orchestrates a hybrid workforce. Their ability to formulate a clear vision and objectives becomes critical, as the autonomous agent will execute the strategy with greater speed and rigor, strictly linked to the definition of the need provided.

The adoption of Artificial Intelligence tools in the product lifecycle is enacting a major transformation. It allows the Product Owner to free themselves from repetitive tasks to focus on highly strategic engagement. Their main challenge shifts:

  • Consolidating the product vision
  • Strengthening ties with stakeholders
  • Assuming greater responsibility for confidential and shared data.

Indeed, by generating analyses or user stories, AI potentially exposes client data and the company’s intellectual property. The PO then becomes the guarantor of the sovereignty of this data and the strict legal compliance (GDPR and AI laws) of the solutions used. This requirement is all the more critical for the consulting PO. Indeed, he must assume an enhanced duty to advise on the risks of leakage.

Let’s consider a concrete example of a Product Owner. Seeking to optimise the testing phase or risk analysis, he might be tempted to copy and paste segments of potentially highly sensitive client data (e.g., addresses, financial information, or transaction history) into a public and unsecured generative AI tool like ChatGPT. This action, performed with the goal of immediate time-saving, exposes this information to unwanted absorption by the AI model.

The usage policy of most of these tools clearly stipulates that submitted data can be used for training the model and its continuous improvement. The risk is, therefore, double and serious in terms of data leakage and regulatory non-compliance, and future reappearance and exploitation. This case perfectly illustrates the necessity for the PO to master data security policies, train their teams on responsible AI usage, and favour the use of enterprise AI tools for handling confidential information, as productivity must never outweigh security and ethics.

More generally, this evolution of the PO towards ethical leadership also gives them the responsibility to detect potential algorithmic biases. This is the case, for example, if the AI suggests a backlog prioritisation that systematically sidelines the needs of a specific population (such as seniors or people with disabilities) simply because they are underrepresented in historical conversion data. The PO thus becomes the essential strategic arbiter facing the complexity of the AI era.

For the Scrum Master: AI as a Facilitator and Coach

The historical guarantor of the process and facilitator of human exchanges, the Scrum Master is also undergoing a complete transformation. According to the analysis conducted by Quanter, the impact of AI on the productivity of this role is one of the highest, with overall gains potentially reaching 50%. Furthermore, Busyqa states that this figure is explained by a 30% decrease in administrative costs and a significant 75% reduction in time spent on meeting documentation.

In addition to facilitating the adoption of Agility, AI is transforming this function, particularly through the following points:

– Project management automation

AI tools can identify potential roadblocks by analysing team tasks, suggest scheduling adjustments, and even summarise sprint progress. The SM no longer has to manually track this information; instead, they receive alerts and recommendations. This frees them up to focus on higher-value tasks, such as coaching the team on more complex issues. A 2024 case study revealed that 86% of Scrum Masters saved between 30 minutes and 2 hours per sprint on reporting thanks to an AI tool (Moza and Ramberg, 2025). These efficiency gains for the Scrum Master have a direct impact on the team. In fact, teams using AI tools recommended by their Scrum Masters reportedly saw a 30% increase in sprint velocity and a 25% reduction in project risks, according to BusyQA (2023).

– Improving retrospectives

AI can analyse the sentiment of team discussions, detect ineffective communication patterns, or identify points of tension. The SM can use this information to direct the retrospective towards underlying issues, such as the challenges of collaborating with the AI itself or managing new skills.

– The challenge of trust and responsibility

AI can suggest a technical solution or a development path. The SM must ensure that the team maintains its ability to make its own decisions and does not blindly rely on the technology’s recommendations. They must encourage critical thinking and shared responsibility within the team.

– The Scrum Master facing autonomous agents

Autonomous agents promise to orchestrate entire agile workflows by detecting bottlenecks in the flow, reassigning resources, or updating the burn-down chart in real-time without human intervention. Facing tools like Devin, capable of managing the lifecycle of a feature from A to Z, the Scrum Master must anticipate the risk of a methodological “black box.” In fact, an AI that optimises the flow too opaquely does not allow the team to master its own process. The Scrum Master then becomes the guarantor of transparency and alignment. He ensures that the automation of workflows serves the team and not the other way around. He’s making sure that the interaction between human developers and the machine remains fluid and non-conflicting.

Transition to a New Role

Based on this analysis, it is clear that AI forces the SM to transition from a role of “ritual guardian” to that of an efficiency coach and a mentor for adaptation to change. Their role is to ensure that the integration of the machine does not dehumanise the agile process. But rather that it augments it intelligently and ethically. However, this challenge includes managing a new “productivity paradox.” While tools like GitHub Copilot accelerate simple coding tasks by up to 55% (Kalliamvakou, 2022), a 2025 study revealed a 19% slowdown for experienced developers working on mature codebases (Becker, Rush, Barnes, and Rein, 2025).

Furthermore, according to a Bain report, actual productivity gains often stagnate between 10 and 15%. These tempered figures highlight an often-overlooked reality on the ground, corresponding to human barriers to adoption. Beyond the technical aspect, the team may face resistance to change. For instance, a steep learning curve for mastering prompting, or even “AI fatigue” due to the incessant evolution of tools. The new mission of the SM (and the PO) is therefore to manage this variance and transform work methods so that AI delivers on its promises. This evolution towards a coaching and self-organisation facilitator role is crucial in the face of AI, as it navigates an environment where automation transforms the very nature of work.

When Humans Remain Irreplaceable

While AI represents a transformative force capable of bringing unprecedented potential for automation and data analysis, it nevertheless faces inherent and fundamental limits that, far from threatening, confirm the importance and sustainability of the Product Owner and Scrum Master roles. Indeed, despite its advances, AI cannot match human intelligence in areas crucial to the success of a product and an agile team. These areas are primarily judgment, empathy, strategic vision, and ethical decision-making.

The Product Owner’s Judgment and Strategic Vision

AI excels at processing historical data and optimising existing processes. But it is incapable of defining an innovative product vision. Or even, anticipating market disruptions, or taking the necessary risks to drive innovation. Therefore, the PO remains essential for:

  • Defining the product vision and strategy. This is a profoundly human task that requires understanding latent needs, interpreting weak signals, and charting a clear direction, even in the absence of formal data.
  • Ethical and contextual prioritisation. While AI can calculate the ROI of tasks, the PO remains the only one who can judge the societal or ethical impact, or the perceived value by the end user, which goes beyond a simple quantitative metric. This involves choosing between complex options where data is incomplete, advising clients on risks related to algorithmic bias (e.g., gender or ethnic bias in prioritisation algorithms) and technological dependence, which could expose them to vulnerabilities in the event of AI failure. Moreover, the PO must preserve team autonomy and encourage critical thinking by integrating diversity to counter the biases inherent in training data.
  • Stakeholder relations. Managing expectations, negotiating, and building consensus with stakeholders are relational and political skills that AI cannot simulate.

The Empathy and Emotional Intelligence of the Scrum Master

The role of the Scrum Master is intrinsically linked to humans and team dynamics, areas where AI shows its most pronounced weaknesses:

  • Coaching and conflict resolution. The SM is at the heart of human interactions. They must demonstrate empathy to understand the tensions, frustrations, or personal barriers that team members face. AI can identify a slowdown, but only the SM can coach the team to resolve the underlying conflict or improve the interpersonal dynamic. They must also raise awareness of risks, such as over-reliance on technology, which can reduce creativity if not managed effectively.
  • Facilitation of continuous improvement. Retrospectives are not just reporting meetings, but safe spaces where emotions and systemic dysfunctions are explored. The SM’s ability to read body language, create a psychologically safe environment, and stimulate self-organisation relies on emotional intelligence, which is inaccessible to algorithms.
  • Maintaining Agile values. The SM is the guardian of Scrum values (courage, focus, commitment, respect, and openness). These values are not rules, but principles of conduct that require nuanced human intervention and personal embodiment.

Conclusion

In conclusion, while AI is becoming a powerful tool for automating the writing of basic user stories, Backlog analysis, or metric tracking (thus freeing up valuable time for POs and SMs), it will never replace the wisdom, intuition, and ethics required for complex decision-making and building high-performing teams. Therefore, the AI era does not signal the end of the PO and SM, but their elevation to more strategic and profoundly human roles.

AI is a powerful assistant that helps optimise processes, but it cannot feel, create, or assume responsibility. The survival of the PO and SM does not lie in their ability to use AI, but in their skill to leverage its strengths while highlighting their own irreplaceable human qualities.

At Devoteam, we believe that the key is not to oppose humans and AI, but to make them co-evolve, because the future of Agility will not depend on the power of machines, but on the wisdom of those who guide them.

Acknowledgements

We would like to thank Zakaria Hzami, Kamal Saib, and Valentin Noël for their valuable suggestions and careful reviews, which helped increase the value and clarity of this article.

Bibliography