How does incorporating Artificial Intelligence (AI) fundamentally transform the responsibilities and skillsets of a Product Owner/Product Manager (PO/PM) within a SAFe® 6.0 environment, and what new practices should they adopt to effectively leverage AI for value delivery?

The integration of Artificial Intelligence (AI) into the Scaled Agile Framework (SAFe®) 6.0 represents a paradigm shift for the Product Owner/Product Manager (PO/PM) role. It evolves the PO/PM from being primarily a steward of the backlog into a strategic, AI-empowered leader who leverages intelligent systems to accelerate value delivery and deepen customer understanding. This transformation impacts core responsibilities, demands new skillsets, and necessitates the adoption of modern practices to harness AI's full potential.

Key Transformations in Core Responsibilities

While the fundamental goals of a PO/PM remain—to maximize the value of the product resulting from the work of the Agile Release Train (ART)—AI augments how these goals are achieved in several critical areas.

• From Manual Research to AI-Driven Customer Centricity

  Traditionally, understanding customer needs involved manual analysis of surveys, focus groups, and feedback channels. An AI-empowered PO/PM uses Natural Language Processing (NLP) and sentiment analysis tools to instantly process thousands of customer reviews, support tickets, and social media comments. This provides a real-time, data-driven view of customer pain points and desires, enabling the PO/PM to build empathy at scale and generate more accurate, impactful feature hypotheses.

• From Intuitive to Predictive Backlog Management

  Prioritizing the ART Backlog using frameworks like Weighted Shortest Job First (WSJF) often involves subjective estimation. AI transforms this by using historical data to predict the effort (Job Size) and potential business value of features. AI can model different sequencing scenarios to forecast their impact on value delivery, helping the PO/PM make more objective, data-informed prioritization decisions. Furthermore, generative AI can assist in drafting initial user stories and acceptance criteria from high-level feature descriptions, significantly reducing administrative overhead.

• Accelerated Ideation and Solutioning

  The PO/PM's role in defining the solution is supercharged by AI. They can use generative AI as a brainstorming partner to explore multiple solution concepts, draft user personas, and create customer journey maps. This co-creation process accelerates the discovery phase, allowing the PO/PM to bring well-formed, innovative ideas to the ART for refinement and implementation much faster than before.

Essential New Skillsets and Practices for the AI-Empowered PO/PM

To effectively operate in this new environment, PO/PMs must cultivate a specific set of skills and adopt new working practices.

• AI Literacy and Prompt Engineering

  A foundational understanding of AI concepts—including machine learning, large language models (LLMs), and their limitations—is non-negotiable. More importantly, the PO/PM must become adept at prompt engineering: the art of crafting precise and effective queries to guide AI tools. This skill is crucial for extracting meaningful insights, generating relevant content, and ensuring the AI's output aligns with strategic goals.

• Data Acumen and Ethical Stewardship

  Since AI models are fueled by data, the PO/PM must develop strong data acumen. This involves understanding data sources, ensuring data quality, and interpreting AI-generated analytics correctly. Critically, the PO/PM becomes a steward for ethical AI. They are responsible for identifying and mitigating potential biases in training data, advocating for transparency in how AI makes decisions, and ensuring that AI-powered features are used responsibly to build trust with customers.

• Continuous Experimentation and Hypothesis Testing

  AI enables a much faster cycle of learning. The AI-empowered PO/PM should champion a culture of continuous experimentation. They can use AI to quickly generate variations of a feature, set up A/B tests, and analyze the results to validate or invalidate hypotheses. This practice shifts the focus from delivering a pre-defined feature to continuously learning and iterating toward the best possible outcome, perfectly aligning with SAFe's Lean-Agile principles.