AI is redefining the way product teams work, says Muhammad Danish

Muhammad Danish explores how AI-driven prototyping, dynamic digital experiences, and evolving governance models are reshaping product roles, workflows, and decision-making

Artificial intelligence is changing more than the speed at which digital products are built. It is challenging assumptions that have shaped product organisations for years: where responsibilities sit and how teams collaborate. 

From handoffs to fluid teams
Traditionally, product managers, designers and engineers have worked through handoffs: one defines the opportunity, one translates it into an experience, one turns that experience into a product. This created specialisation, but also delays and misunderstandings. 

AI is making those boundaries more permeable. Product managers can create functional prototypes. Designers can work directly with code. Engineers are increasingly involved in design decisions, exploring user experiences and interface concepts with far less effort than before. 

AI is blurring traditional roles. Boris Cherny, creator of Claude Code at Anthropic, offers a compelling example: as engineering, product, design, and data science melt into one kind of role, what emerges is a set of archetypes rather than job titles. Prototypers generate ideas, most of which never ship. Builders turn prototypes into production-grade products. Sweepers simplify the system, unship what is not earning its place, and optimise performance. Growers iterate on what has shipped to improve product-market fit. Maintainers keep a mature system reliable as it scales. 

These archetypes are not tied to job function: across Anthropic, some designers fit the first category, some engineers the third, and many span two or three. Specialist expertise will not disappear, but AI allows specialists to operate far outside traditional boundaries. The question is not which function someone belongs to, but what they can contribute. 

As execution becomes easier, judgment becomes more valuable. AI will not tell a team which problem deserves attention, or what should never be built. 

From fixed journeys to dynamic experiences
Now imagine that predefined journeys no longer exist. You want to renew your driving licence or apply for a home loan in the UAE. Today, every applicant moves through the same fixed sequence of screens. Instead, imagine the portal generating the next step as it goes, based on what it already knows about you: your eligibility, your history, your permissions. No two users would see the same interface. 

No platform in the UAE works this way today. It is a thought experiment rather than a case study, but a useful one: it describes a plausible destination for digital services. 

This changes the product challenge. Teams are no longer designing a fixed sequence of screens. They are designing an intelligent service that must understand intent, respect permissions, work across systems, and remain accountable. The interface becomes an outcome of that understanding, not the starting point. Designers set principles for trust and human control; product managers set outcomes; engineers focus on orchestration. That changes not only how products are built, but what a digital product can be. 

Organisations should therefore avoid treating AI as a way to accelerate every existing process. If it only produces requirements faster or moves work between the same approval stages, companies become faster without becoming better. The greater opportunity is to ask whether some handoffs, documents and approval points are still necessary, and to redesign workflows around faster learning and better decisions. 

The enterprise governance test
The greatest challenge will be organisational rather than technical. Allowing designers to build or product managers to prototype raises questions about access controls, security, compliance and accountability.

Once a prototype begins to influence customers or operations, the distinction between experimentation and production becomes critical. 

There are equally important questions about quality and careers: how to protect design standards when interfaces can be generated by people without design training, how to assign responsibility when several disciplines contribute, and how career progression works when expertise is demonstrated through capabilities. 

These questions cannot be answered by technology alone. The differentiator will be which organisations develop the right operating model: policies, workflows and governance that enable speed without sacrificing control. Enterprises that resolve this early will learn to distribute software creation safely while others are still deciding where responsibility sits. 

Prioritisation as advantage
The product leaders who create the most value will not be those who produce the most output. They will frame problems accurately and tell an impressive story from a product that solves a real customer need. When execution is abundant, taste and prioritisation become strategic capabilities. 

AI is not making product roles disappear, but it is making their boundaries more flexible. Organisations that understand this earliest will build products that are faster, more relevant and more responsible. Every product organisation should ask one uncomfortable question: if it were designed today, with AI from the beginning, would it look the same? 

This opinion piece is authored by Muhammad Danish, Senior Director, Product Design, Corporate & Institutional Banking, Emirates NBD. The views expressed in this article are those of the author and do not necessarily reflect the views of his employer.