AI in product management is no longer a question of if. It is a when. And when we say 'when,' we are not talking about years. We are talking months, given the pace of innovation and adoption.

A new study in Management Review Quarterly, “Where does AI play a major role in the new product development and product management process?” by Aron Witkowski and Andrzej Wodecki, maps out the current state of AI in product work. It synthesizes more than 190 publications and practitioner insights to show where AI is already embedded and where research has yet to catch up.

Where AI is Already Making an Impact

The research, combined with our daily experience, highlights early discovery as one of the most active areas of AI.

Product teams are already experimenting with AI-powered tools that generate product requirements documents, saving hours of manual formatting and ensuring consistency across teams.

Some PMs use AI-driven prototyping platforms that can translate a natural language prompt into a clickable interface, allowing faster validation of early ideas. Others lean on AI assistants that generate user stories directly from interview transcripts or customer feedback, making backlog grooming less about rewriting and more about prioritizing.

These examples show how AI is no longer limited to analytics or personalization. It is working its way into the daily fabric of product management—the unglamorous but essential tasks that keep product cycles moving.

Where Research is Still Lacking

The Witkowski study also identifies blind spots. There is a lack of systematic research on how AI can influence concept testing or post-launch validation. Integrative frameworks that span the entire product lifecycle are almost absent. Questions of trust and transparency, which directly affect adoption, remain underexplored.

This gap matters. The activities most critical to practitioners—how AI can accelerate experiments or help close the loop after release—are barely documented in academic work. For example, while AI can now generate dozens of prototype variants instantly, there is little evidence to suggest whether this accelerates decision-making or overwhelms teams with choices. Similarly, while AI can generate user stories, there is no structured evidence that these outputs improve product outcomes or prevent the introduction of new biases.

Why This Matters for PMs

For PMs, the takeaway is that you are the testbed. The tools exist, but the validated playbooks do not. Your experiments with PRD generators, your trials of vibe-coding prototypes, and your attempts to scale user stories with AI are not just tactical choices—they are contributions to the collective understanding of what works.

Because academic research often lags by years, product managers are often the early researchers in their own field. Each team’s learnings can ripple outward, closing the gap between practice and theory.

Closing Thought

AI is already woven into the toolkit of product managers, and its presence will only grow. The question is not whether it belongs, but how quickly PMs can adopt responsibly and share what they learn. By using today’s tools thoughtfully and documenting outcomes, product managers do not just adapt to the future—they help define it.

Claude for Chrome (now in pilot), Perplexity’s Comet, and Dia are all pushing the idea of a browser that doesn’t just display pages but acts within them. But as soon as you let an AI click, type, and execute, the hardest problem comes into view: security.

The quiet threat of prompt injection

Anthropic deserves credit for going deep on vulnerabilities in its Claude for Chrome pilot.

“Some vulnerabilities remain to be fixed before we can make Claude for Chrome generally available. Just as people encounter phishing attempts in their inboxes, browser-using AIs face prompt injection attacks—where malicious actors hide instructions in websites, emails, or documents to trick AIs into harmful actions without users’ knowledge.”

In red-team testing, Claude executed malicious instructions in nearly one in four targeted cases. With defenses such as site restrictions, confirmations, and domain blocking, the rate dropped to 11 percent. For browser-specific tricks like hidden form fields, the success rate fell from 35 percent to zero.

Publishing these numbers is significant. Transparency matters when the attack surface is any page on the internet.

The optimistic angle

This is not bleak news. First, we now have metrics—attack success rates, challenge sets, benchmarked defenses—that create a shared language for researchers, vendors, and enterprises. Second, mitigations are already in play: permission scaffolding, domain filters, and classifiers are being tested in live pilots. Third, vendors like Anthropic are treating security as a design pillar, not an afterthought—limiting early releases to small groups, requiring explicit consent for risky actions, and publishing residual risks.

Security is not a side concern for agentic browsers. It is the defining challenge that will determine whether they are trusted or discarded. The encouraging sign is that, unlike earlier waves of technology, this one is starting with adversarial testing and layered defenses built in.

The headwinds are real. But so is the progress. If the future of browsing is agentic, the real race is not who ships first—it is who builds the browser that can act safely.

The most important product decision you make is not the roadmap. It’s not the features you prioritize or the markets you enter. It is the culture you build.

Culture is not a poster on the wall or a slide in a town hall. It is how decisions get made when nobody is looking. It is how teams respond to setbacks, how they argue about priorities, how they treat customers when tradeoffs get hard. Culture is the invisible operating system of your product organization — and it shapes every outcome.

Culture Is the Real Product

Features will evolve. Technology will shift. Markets will turn. What endures is the way your teams think, decide, and collaborate. Marty Cagan has argued that great product companies don’t win because they have better ideas. They win because they have empowered cultures where teams are trusted to make decisions close to the customer.

If you want consistent product excellence, you cannot rely on strategy slides or innovation slogans. You need a culture that sustains curiosity, collaboration, and courage every single day.

Everyday Behaviors Define Culture

Culture is not abstract. It lives in small, repeated actions.

• Do roadmap discussions encourage dissent or punish it?
• Do teams treat customer feedback as a gift or as an interruption?
• Is failure treated as a data point or as a career risk?

People crave belonging, but belonging is not the same as fitting in. The strongest cultures are those where people feel safe bringing their perspectives without needing to mimic everyone else. In product teams, that balance — belonging without conformity — is what drives both speed and innovation.

Hiring and Onboarding as Culture Design

Hiring is one of the clearest ways leaders shape culture, but it is only the beginning. When you bring someone into the team, you’re not just adding skills. You’re reinforcing or eroding cultural norms.

The goal is not to hire people who simply “fit in.” That creates mirrors, not complements. The goal is to hire people who align with our core values — including curiosity about customers, comfort with ambiguity, and a willingness to collaborate — while adding new perspectives that stretch the team’s thinking.

Equally important is how you onboard. Even the best hires will drift if your rituals and norms don’t reinforce the culture you want. The way teams run standups, review designs, or handle retrospectives all shape how new people understand “how we do things here.”

Leaders as Culture Carriers

Leaders cannot delegate culture to HR or an annual workshop. They are culture carriers. Teams watch what leaders reward, what they tolerate, and what they model.

A leader who celebrates data-driven insights but ignores them in decision-making weakens the culture. A leader who claims collaboration is a value but rewards only individual heroics undermines the value of collaboration. Conversely, when leaders empower teams, welcome dissent, and stay customer-focused, those behaviors cascade across the organization.

Strong product cultures strike a balance: aligned enough to move fast, diverse enough to avoid blind spots. Belonging gives teams cohesion. Stretch gives them creativity.

Always Dynamic

Culture is not static. It is designed, reinforced, and tested every day. Hiring shapes it. Onboarding cements it. Rituals sustain it. Leadership amplifies it.

As a product leader, your roadmap matters. But the culture behind it matters more. Because in the end, the features your teams ship are built on the foundation of who they are and how they work together. Culture is the product you ship every single day.

Key Takeaway

The jury is still out on whether AI will replace or empower software developers, but dismissing junior talent is a short-sighted approach. Their curiosity and adaptability make them the best positioned to thrive in an AI-driven future—qualities that matter more than years of experience.

Why This Matters

AI is reshaping the nature of engineering work. Leaders face pressure to cut costs and experiment with automation. Some argue junior developers are the most “replaceable” role. Others counter that they are the most essential. The question is not simply about efficiency, but about who will carry engineering organizations into the next era.

Two Diverging Perspectives

The Replacement View

Several leaders predict AI will soon take over tasks typically assigned to entry-level engineers. Dario Amodei believes AI could generate up to 90 percent of new code. Sam Altman has said jobs will “definitely go away.” Geoffrey Hinton has gone further, warning that AI could eventually replace white-collar work broadly.

This argument positions juniors as expendable—valuable tasks automated, oversight left to senior staff.

The Augmentation View

Others view AI as a lever, not a replacement. Thomas Dohmke expects the smartest companies to hire more engineers, not fewer. Andrew Bosworth describes AI as expanding the capabilities of developers. Mustafa Suleyman warns more about a widening skills gap than job loss.

This view highlights a crucial fact: juniors bring curiosity and drive that seniors, conditioned by legacy practices, often lack. As AWS CEO Matt Garman noted in a podcast (video), they are the most eager to experiment with new tools. Curiosity is not a “soft” trait—it’s the core ingredient for mastering rapidly evolving technologies.

Why Curiosity Beats Tenure

Years of experience are not always a proxy for adaptability. In many cases, “20 years of experience” can mean repeating the same year’s practices twenty times. Senior developers carry the weight of how things used to be done. Juniors, by contrast, arrive without baggage, ready to adopt AI workflows, test new approaches, and ask questions that break conventional thinking.

It's not even junior or senior: are you intellectually curious, hungry to make a difference, and ready to reinvent yourself at work?

This difference matters because AI is not just a coding assistant—it’s a paradigm shift. The engineers most willing to learn, unlearn, and relearn will be the ones who define the next wave of software.

Takeaway for Product and Tech Leaders

The debate is unresolved, but the path forward is clear:

• Value curiosity as much as expertise. Juniors bring energy and openness that AI will reward.
• Build a dual strategy. Utilize AI to automate repetitive tasks while investing in mentoring early-career engineers.
• Avoid false efficiency. Cutting junior roles may deliver short-term savings, but risks hollowing out the future talent pipeline.

Conclusion

AI may change how code is written, but curiosity, adaptability, and drive are timeless assets. Junior developers embody these traits more than any other group. The leaders who cultivate them—rather than replace them—will build organizations ready for whatever this fast-moving future holds.

APIs are often seen as back-office plumbing, but in the emerging world of agentic AI, they are the execution layer that makes autonomy possible. Without APIs, AI remains stranded in theory—able to reason, but unable to act.

From Copilots to Agents

The last wave of AI adoption has been copilots—tools that help users write emails, summarize documents, or draft code. These copilots assist, but they don’t take initiative. Agentic AI is different. Agents can plan, reason, and execute tasks end-to-end, often without human intervention.

But here’s the catch: for agents to actually do something, they need APIs. APIs are the hands and feet of AI. They let an agent check a customer’s eligibility, process a payment, or reschedule a shipment.

APIs as the Action Layer

Consider healthcare. A patient-facing AI agent may be asked to verify benefits and book an appointment. That agent cannot achieve the goal without a secure and reliable API. Through APIs, it connects to claims systems, checks eligibility, and schedules care—all actions that today require phone calls and manual lookups.

Let me take an example close to my heart (I work at Optum). At Optum, APIs like those available on the Developer Portal already provide secure access to eligibility, claims, and clinical functions. This means agentic AI can be layered on top to handle tasks that previously bogged down patients, providers, and administrators.

The same story plays out across industries:

• A fintech agent creates an invoice through Stripe’s APIs.
• A retail agent adjusts staffing using HR and scheduling APIs.
• A logistics agent reroutes shipments by orchestrating supply chain APIs.

In every case, APIs are the indispensable bridge between an AI’s reasoning and real-world outcomes.

Why This Matters for Leaders

Leaders who think of APIs as minor technical connectors risk missing the bigger shift. As agentic AI moves from hype to reality, APIs will determine whether your organization can harness it to deliver value. If your APIs are robust, secure, and well-documented, they become the foundation for intelligent automation and new business models. If they’re neglected, your AI strategy stalls at the whiteboard.

Takeaway: Agentic AI is only as powerful as the APIs it can call. Leaders should treat APIs as strategic assets—the execution layer that will determine whether AI can move from promising demos to meaningful business outcomes.

The claim that there are no new ideas in AI, only new datasets, is both provocative and partly true. As Jack Morris argued in his recent post, many of the most important AI milestones have been driven not by theoretical leaps, but by new sources of data.

He puts it succinctly:

“The breakthroughs weren’t big ideas; they were new ways to learn from new kinds of data.” from blog.jxmo.io

That framing resonates with history. The ImageNet dataset unlocked computer vision. Web-scale text collections made Transformers viable. Reinforcement learning from human feedback reshaped how models align with our preferences. Time and again, progress has come from scaling access to structured, abundant inputs rather than from fundamentally novel algorithms.

But the story is incomplete if we stop there. Innovation in AI is multidimensional, and algorithms still matter. DeepMind’s AlphaEvolve has already demonstrated the ability to generate algorithms beyond human expertise, producing approaches to matrix multiplication and optimization problems that surprised even expert researchers, as reported in Wired. If data were the only driver, we would not see systems out-innovating decades of human design.

As Dr. Rumman Chowdhury cautions:

“Innovation stems from human minds, not AI. Don’t delegate your thinking to machines.” from Business Insider

Her warning underscores the deeper point: humans frame the problems, interpret the outputs, and decide which ideas matter. AI extends our reach, but it does not replace human originality. Even when AI generates unexpected solutions, the spark of innovation is in how people direct, interpret, and apply them.

Morris’s post highlights an important truth: data has been the engine of AI’s rise. But the full picture includes algorithms, architectures, and above all, human creativity. Ignoring these other drivers risks oversimplifying how innovation actually happens.