By Markus Nispel, CTO EMEA and Head of AI Engineering, Extreme Networks
AI Appreciation Day quickly came and went in July, and much of the conversation still focused on its potential. My view is that companies need to be prioritizing the next steps – the time for experimentation is over, and AI investments are under more scrutiny than ever. As we move forward, organizations should be focusing on the practical ways AI is transforming everyday responsibilities for IT individuals, and the mindset shift that will come with this automated era.
The biggest shift isn’t that AI can complete individual tasks more quickly. It changes how entire workflows function. It enables people to move from research to execution faster, uncover insights that would otherwise take significantly longer to find, and spend less time on repetitive, manual work and more time applying their expertise.
Within the next year, 57% of IT leaders plan to remove humans from AI decisioning. However, the technical teams that will see the greatest value won’t be those that fully automate processes. They’ll be the ones that securely use it to strengthen capability, improve decision making and create capacity for the creative, complex work that people do best.
Rethinking the Way We Innovate
AI is transforming leadership by giving leaders better data, deeper insight, and greater transparency. Personally, it has changed every stage of my work from research to execution. It helps me identify risks, keep initiatives on track and, when used correctly, can accelerate the journey from an idea to a demonstratable prototype. It’s a fundamentally different way of innovating.
While it does not invent anything by itself, it does supercharge human creativity. It doesn’t just answer questions but challenges assumptions, connects ideas, and helps turn early thinking into something worth building.
Making Time for Higher-Value Work
Another big advantage is how quickly it can synthesize information. For example, company assessments for partnership and M&A considerations that once took days can now be completed in hours, freeing me to focus on strategy, judgment and where I can add the most value.
AI won’t change every role in the same way, but it consistently frees people from time-consuming tasks so they can prioritize this higher-value work. We’re already seeing experienced teams use AI to deliver significantly greater productivity (particularly in software development) while improving both quality and speed.
The Human Role Isn’t Going Away
The greatest opportunity isn’t replacing human expertise; it’s amplifying it. While AI can inform decisions, it can’t own them. Leadership will always be about judgment, accountability and the courage to make the final call.
AI is becoming more capable, and organizations need to think carefully about what it really means to keep a human in the loop. For a long time, the answer has been simple: assess the output and approve it to preserve accountability. Yet as systems become more reliant on AI and the volume of automated decisions increases, approval is gradually becoming a routine rather than a conscious decision. It becomes a click instead of a choice.
That’s where organizations need to be careful. Leadership isn’t about signing off on automation; it’s about knowing when to question it, when to intervene and ultimately take responsibility for the outcome. Technology should strengthen human judgment, not distance people from it.
The same principle applies to governance. As autonomous agents begin querying systems, making configuration changes and supporting critical workflows, they effectively become users in their own right. They need clear identities, defined permissions and appropriate oversight, just like human operators. Organizations should have visibility not only into what an agent did, but what it was permitted to do and why. Building trust starts with confidence that these systems are operating within clearly defined boundaries.
In recent weeks, reports have emerged of AI agents overstepping their roles and expanding beyond their specific assignments. While this is creating some wariness around agents, the underlying issue is whether the infrastructure is properly built to keep agents within the proper boundaries. Security boundaries must be created and governed outside of AI’s reasoning, so agents cannot push past them without human approval. When AI oversteps, it’s ultimately the people who deployed it who are responsible.
Agentic decisions will be a force multiplier for efficiency, but people will continue to own the outcomes. With strict control parameters and human oversight, AI can be the superpower it promises to be – and already is for some of us.
