Agile in an AI-driven world: What’s really changing
Agile has long been treated as a fixed concept. Teams adopted certain rituals, like stand‑ups, sprints and backlogs, and organisations assumed this was enough to be “agile”.
If this was ever true, it’s not anymore. AI has reshaped the way we work, and our approaches must keep up to the new reality. Agile hasn’t disappeared though; it’s just evolving.
So, we brought together a panel of experts at our recent Beyond Agile event, to unpack what’s changing, the role that agile really plays in 2026, and how to get the most out of it. Here’s what they revealed.
Why agile still matters, but not as we knew it
Panellist Lee Dunn, Head of the Digital Academy for Scottish government, captured it perfectly; “the term is still useful, but the meaning behind it has changed.”
Agile is no longer defined by daily ceremonies or fixed frameworks. What matters is how organisations can adapt, learn and respond in an environment where information, tools and possibilities are shifting fast.
The core principles of agile remain relevant: break work into manageable pieces, stay close to customers, iterate, reduce risk through learning. But the execution looks very different today.
As Imran Ahmad, Senior Advisor for Agile and Scrum at PeopleCert, put it, “agile was always about adapting faster.”
With AI accelerating delivery, organisations must rethink how teams make decisions, prioritise work and understand value. The shift isn’t away from agile – it’s towards a more mature, product‑led interpretation of it.
AI is reshaping delivery, and raising expectations
Our panel explored three big changes AI brings to modern delivery teams:
1. AI demands better clarity
Teams can no longer rely on vague estimations or loosely defined tasks. AI runs best on detailed, well‑structured input, pushing teams to define intent and outcomes with far greater precision. Agile working can still support this, by ensuring that the clarity and the due process to enable quality AI outputs are in place.
2. Workflow automation accelerates everything
AI now structures documents, rewrites content, generates code and analyses data instantly. For example, enabling us to save time by “talking” out an email and have it instantly rewritten in the correct structure – as Brett StClair, CoFounder and Chief Connector at Teraflow.ai, highlighted.
But, while AI reduces time spent typing or structuring content, it increases the need for teams to apply judgment, reflect on meaning and ensure that rapid output aligns with product goals.
This reduction in time spent doing, increases the opportunity (and the need) for teams to spend time thinking – interpreting, critiquing and validating AI‑generated work. The emphasis on human skills and agility of thinking is compounded.
3. Trust becomes a new agile principle
A new model of AI-human collaboration is emerging:
- Phase 1: human initiates → AI works → human validates
- Phase 2: AI initiates → AI works → human validates
- Phase 3: AI initiates → AI works → AI validates
To be clear, this isn’t about phasing the human element out. This evolution will enable never-before-possible processes and innovation, but it can only happen through reliance on human input.
In other words, you must start at phase 1. You must teach AI how to do good work, and so you need good people to teach it.
AI maturity will be built on trust through experience, governance and human judgement. Temptation to skip these stages will lead to risk, quality decline, wasted time, lost ROI... the list goes on.
Ruth Mbegabolawe, Senior Business Analyst at Aegon, highlighted that AI should help create more space for this kind of meaningful product thinking. “AI will give us more time to think about why we’re doing this,” Ruth said, emphasising that AI enables teams to deepen their understanding of stakeholder needs and customer outcomes.
The shift is about refocusing on intent: questioning why products are being built, why changes are made, and who will ultimately benefit.
The real risk: losing foundational skills
A key concern right now is that rapid adoption of AI can mask a lack of foundational understanding.
Jamie Reed, Product Owner at Zudu, cautioned that some of the foundational practices in agile – like writing user stories – are more than administrative tasks. “You put yourself in the shoes of a user… there’s a risk we become flabby if we let the tools master us,” he said. Jamie explained that even the socalled “donkey work” has value, because it builds empathy, sharpens thinking and can even spark creativity. If teams outsource this entirely to AI, they risk weakening the very skills that make products meaningful.
Early-career professionals may produce impressive output quickly – in Brett’s words, “They don’t check it, they don’t validate it… they rush to get it into production.”
But these skills will be needed for the future, to ensure continued quality assurance and competent oversight.
AI amplifies skill gaps as quickly as it multiplies capability. Therein lies the real problem statement when it comes to AI, talent and work.
So, what can organisations do about it?
Why training matters more than ever
To thrive in an AI-driven world, organisations can’t rely on frameworks alone.
- Teams need skills in:
- Critical thinking
- Product‑led decision‑making
- Writing and structuring clear requirements
- Validating AI outputs
- Understanding delivery fundamentals
- Adapting confidently amidst ambiguity
So, has AI made agile irrelevant?
On the contrary. In the AI age, human capabilities are your new competitive advantage. Agility has never been more critical.
If you’re ready to get ahead with AI, start with skills. Talk to QA.
