Ho Wai CHENGProgram Manager| BankHong Kong, Hong Kong
AI is increasingly embedded in Agile project environments. Teams are using it to draft sprint goals, summarize retrospectives, analyze delivery data, and support planning and forecasting. On the surface, this looks like a clear acceleration of Agile delivery. But speed doesn’t always equal progress.
Without clear intent and leadership oversight, AI can reinforce existing biases, obscure accountability, or create a false sense of confidence in plans and forecasts. Agile has always emphasized learning, transparency, and human collaboration—principles that can be strengthened by AI, or quietly undermined by it.
For project and PMO leaders, the real question isn’t whether AI should be used, but how it is shaping decisions, behaviors, and outcomes. Are teams using AI to deepen insight and improve judgment—or to bypass critical thinking and conversations?
AI has the potential to elevate Agile delivery, improve predictability, and free teams to focus on value. But only when it’s applied intentionally, ethically, and in service of people—not just process efficiency.
Curious to hear from others:
Where has AI genuinely improved Agile delivery on your projects?
Where has it fallen short or created new challenges?
What guardrails or principles guide AI use in your teams today?
Looking forward to learning from real‑world experiences.
Saving Changes...
Sort By:
Luis BrancoCEO| Business Insight, Consultores de Gestão, LdªCarcavelos, Lisboa, Portugal
This is a timely and well-framed question. What’s at stake is not Agile speed, but the quality of sense-making behind that speed.
In many teams, AI is already accelerating artefacts – sprint goals, reports, forecasts. That part is real. The illusion appears when acceleration is confused with progress, and outputs are mistaken for understanding.
From what I see in mature delivery environments, AI creates value when it strengthens three things, not when it bypasses them: judgment, dialogue, and accountability. Pattern detection, scenario exploration, and cognitive load reduction are legitimate gains. Delegating thinking, responsibility, or ethical ownership is not.
Agile was never just about faster cycles. It was about shortening feedback loops so humans could learn better together. If AI weakens those learning conversations, even subtly, predictability may increase while wisdom decreases.
For leaders and PMOs, the real governance question is therefore simple but demanding: Are we using AI to enhance collective intelligence, or to mask unresolved tensions, trade-offs, and decision ownership?
The guardrails I see working best are human, not technical:
– explicit ownership of decisions, regardless of AI input – deliberate pauses for sense-checking and challenge – transparency about where AI informs, and where humans decide
When applied with intent, AI can deepen Agile maturity. Without that intent, it risks becoming a very efficient way of moving faster in the wrong direction.
Looking forward to learning from concrete experiences where AI truly elevated judgment, not just velocity.
...
1 reply by Ho Wai CHENG
Feb 02, 2026 10:25 AM
Ho Wai CHENG
...
Thank you for the thoughtful perspective. Your distinction between accelerating artefacts and improving sense‑making is an important one and resonates strongly.
I agree that AI creates value only when it strengthens judgment, dialogue, and accountability—not when acceleration is mistaken for progress. Your reminder that Agile is about shortening learning loops, not just cycles, is especially relevant. I’ve seen predictability improve through AI‑supported insights while the quality of planning and retrospective conversations declined due to insufficient challenge.
Your emphasis on human guardrails—clear decision ownership, deliberate pauses for sense‑checking, and transparency about how AI informs decisions—captures what’s needed to ensure AI enhances collective intelligence rather than obscuring trade‑offs.
Looking forward to learning from practical examples where teams have struck this balance well.
Saving Changes...
Sergio Luis ConteHelping to create solutions for everyone| Worldwide based OrganizationsBuenos Aires, Argentina
First of all, AI is a board term. AI is embedded in any type of approach, agile for example, from more than 30 years ago. Ai is just a tool, as any other type a tool. No more than that. In this case a tools that helps all of us to analyze a great amount of data. In the last times, from 2017, generative AI is what causes a disruption. So, AI will not improve delivery, acceleration, etc etc just in agile approach. AI is doing that in any type of initiative to create a solution. People that dont understand is giving the first step to fail.
...
1 reply by Ho Wai CHENG
Feb 01, 2026 11:11 AM
Ho Wai CHENG
...
Thank you for sharing your perspective. I agree that AI, at its core, is a tool—and like any tool, its value depends on how well it is understood and applied. Data analysis, automation, and decision support have existed in Agile environments for decades in different forms.
Where I see in recognizing that generative AI introduces a different scale and speed of influence, particularly in how teams plan, forecast, and make decisions. Without clarity of purpose, governance, and human judgment, AI can just as easily reinforce poor assumptions as it can improve insight.
So rather than positioning AI as either a breakthrough or a non-factor, I see it as an amplifier—of both strengths and weaknesses in Agile practices. The real risk isn’t using AI, but using it without sufficient understanding and transparency.
In that sense, I agree with your point: misunderstanding the tool is often the first step toward failure. The challenge for Agile leaders today is ensuring AI supports critical thinking and collaboration, rather than replacing them.
Saving Changes...
Ho Wai CHENGProgram Manager| BankHong Kong, Hong Kong
Feb 01, 2026 10:11 AM
Replying to Sergio Luis Conte
...
First of all, AI is a board term. AI is embedded in any type of approach, agile for example, from more than 30 years ago. Ai is just a tool, as any other type a tool. No more than that. In this case a tools that helps all of us to analyze a great amount of data. In the last times, from 2017, generative AI is what causes a disruption. So, AI will not improve delivery, acceleration, etc etc just in agile approach. AI is doing that in any type of initiative to create a solution. People that dont understand is giving the first step to fail.
Thank you for sharing your perspective. I agree that AI, at its core, is a tool—and like any tool, its value depends on how well it is understood and applied. Data analysis, automation, and decision support have existed in Agile environments for decades in different forms.
Where I see in recognizing that generative AI introduces a different scale and speed of influence, particularly in how teams plan, forecast, and make decisions. Without clarity of purpose, governance, and human judgment, AI can just as easily reinforce poor assumptions as it can improve insight.
So rather than positioning AI as either a breakthrough or a non-factor, I see it as an amplifier—of both strengths and weaknesses in Agile practices. The real risk isn’t using AI, but using it without sufficient understanding and transparency.
In that sense, I agree with your point: misunderstanding the tool is often the first step toward failure. The challenge for Agile leaders today is ensuring AI supports critical thinking and collaboration, rather than replacing them. Saving Changes...
Ho Wai CHENGProgram Manager| BankHong Kong, Hong Kong
Jan 31, 2026 10:09 AM
Replying to Luis Branco
...
This is a timely and well-framed question. What’s at stake is not Agile speed, but the quality of sense-making behind that speed.
In many teams, AI is already accelerating artefacts – sprint goals, reports, forecasts. That part is real. The illusion appears when acceleration is confused with progress, and outputs are mistaken for understanding.
From what I see in mature delivery environments, AI creates value when it strengthens three things, not when it bypasses them: judgment, dialogue, and accountability. Pattern detection, scenario exploration, and cognitive load reduction are legitimate gains. Delegating thinking, responsibility, or ethical ownership is not.
Agile was never just about faster cycles. It was about shortening feedback loops so humans could learn better together. If AI weakens those learning conversations, even subtly, predictability may increase while wisdom decreases.
For leaders and PMOs, the real governance question is therefore simple but demanding: Are we using AI to enhance collective intelligence, or to mask unresolved tensions, trade-offs, and decision ownership?
The guardrails I see working best are human, not technical:
– explicit ownership of decisions, regardless of AI input – deliberate pauses for sense-checking and challenge – transparency about where AI informs, and where humans decide
When applied with intent, AI can deepen Agile maturity. Without that intent, it risks becoming a very efficient way of moving faster in the wrong direction.
Looking forward to learning from concrete experiences where AI truly elevated judgment, not just velocity.
Thank you for the thoughtful perspective. Your distinction between accelerating artefacts and improving sense‑making is an important one and resonates strongly.
I agree that AI creates value only when it strengthens judgment, dialogue, and accountability—not when acceleration is mistaken for progress. Your reminder that Agile is about shortening learning loops, not just cycles, is especially relevant. I’ve seen predictability improve through AI‑supported insights while the quality of planning and retrospective conversations declined due to insufficient challenge.
Your emphasis on human guardrails—clear decision ownership, deliberate pauses for sense‑checking, and transparency about how AI informs decisions—captures what’s needed to ensure AI enhances collective intelligence rather than obscuring trade‑offs.
Looking forward to learning from practical examples where teams have struck this balance well. Saving Changes...
Program Manager| HARPER SRLSanto Domingo / Distrito Nacional, Dominican Republic
AI help when it reduces cognitive load and surfaces patterns teams can discuss together. It falls short when outputs replace conversation and ownership gets fuzzy. The guardrail is simple: AI can inform, but decisions, and their consequences, always stay human.
...
1 reply by Ho Wai CHENG
Apr 06, 2026 4:38 AM
Ho Wai CHENG
...
I agree that AI adds the most value when it reduces cognitive load and highlights patterns that teams can interpret together. That framing keeps collaboration at the center of Agile rather than replacing it.
Your guardrail is powerful in its simplicity: AI can inform, but ownership of decisions must remain human. When outputs start substituting for dialogue, that’s usually where accountability and shared understanding begin to erode. Keeping AI as an input to thinking — not a replacement for it — feels like the right balance.
On multiple teams, AI genuinely helped us move faster - it summarized retros, spotted delivery patterns, and reduced admin work so we could focus on real problem-solving. That felt like true acceleration.
But I’ve also seen AI-generated forecasts look impressively confident while hiding weak assumptions underneath. The plans looked smarter than the thinking behind them. That’s where it becomes an illusion.
...
1 reply by Ho Wai CHENG
Apr 06, 2026 4:41 AM
Ho Wai CHENG
...
I’ve seen the same pattern: when AI reduces admin work and surfaces delivery insights, it genuinely creates space for deeper problem-solving. That kind of acceleration feels real because it improves focus and clarity.
AI can present outputs with a level of polish and confidence that masks fragile assumptions underneath. When the plan looks smarter than the thinking behind it, that’s exactly where illusion replaces insight. It reinforces the need to interrogate assumptions and make the reasoning visible — not just accept the output because it appears sophisticated.
Saving Changes...
Ho Wai CHENGProgram Manager| BankHong Kong, Hong Kong
Feb 07, 2026 9:49 PM
Replying to Lissette Indhira Pimentel Sosa
...
AI help when it reduces cognitive load and surfaces patterns teams can discuss together. It falls short when outputs replace conversation and ownership gets fuzzy. The guardrail is simple: AI can inform, but decisions, and their consequences, always stay human.
I agree that AI adds the most value when it reduces cognitive load and highlights patterns that teams can interpret together. That framing keeps collaboration at the center of Agile rather than replacing it.
Your guardrail is powerful in its simplicity: AI can inform, but ownership of decisions must remain human. When outputs start substituting for dialogue, that’s usually where accountability and shared understanding begin to erode. Keeping AI as an input to thinking — not a replacement for it — feels like the right balance.
Saving Changes...
Ho Wai CHENGProgram Manager| BankHong Kong, Hong Kong
Feb 22, 2026 3:27 PM
Replying to Bruce Buryo
...
On multiple teams, AI genuinely helped us move faster - it summarized retros, spotted delivery patterns, and reduced admin work so we could focus on real problem-solving. That felt like true acceleration.
But I’ve also seen AI-generated forecasts look impressively confident while hiding weak assumptions underneath. The plans looked smarter than the thinking behind them. That’s where it becomes an illusion.
I’ve seen the same pattern: when AI reduces admin work and surfaces delivery insights, it genuinely creates space for deeper problem-solving. That kind of acceleration feels real because it improves focus and clarity.
AI can present outputs with a level of polish and confidence that masks fragile assumptions underneath. When the plan looks smarter than the thinking behind it, that’s exactly where illusion replaces insight. It reinforces the need to interrogate assumptions and make the reasoning visible — not just accept the output because it appears sophisticated. Saving Changes...
Community Champion
