Digital Economy Dispatch #248 -- The Summer's Over -- Time for an AI Strategy Reset!

I've always had a complicated relationship with September. As a child, I dreaded that familiar knot in my stomach as summer wound down – the one that came with buying new notebooks, sharpening pencils, and mentally preparing for another year of early mornings and homework. It’s a feeling that continued long after my school days were over. Yet looking back, I realize those back-to-school moments were crucial reset points. They forced me to take stock, establish new routines, and approach challenges with fresh perspectives.

After taking some time away this summer to recharge, I'm struck by how this same back-to-school energy might be exactly what we need right now as we all return to work and face increasing pressures to adopt new AI technologies. September feels like the perfect time to pause, reassess our AI strategies, and, most importantly, get brutally honest about whether we're actually prepared for the transformation ahead.

It’s a reset that is desperately needed. While many of us have been taking a break, an uncomfortable truth has been surfacing: For most organizations, AI adoption isn’t going very well. Although everyone's talking about AI's potential, too many organizations have been rushing toward implementation without addressing the foundational realities that will determine success or failure. It's time for some honest conversations with your teams about three critical areas that can't be glossed over anymore.

Technical Infrastructure: The Foundation of AI Delivery

The starting point is to review the real state of your technical infrastructure. I've been in too many meetings where executives get excited about AI capabilities while their IT teams quietly panic about systems that were already struggling before AI entered the picture. Now all they see is additional pressure on their crumbling infrastructure and understaffed IT teams.

Be honest with yourself and your team: Can your current infrastructure actually support the added requirements of AI workloads? We're not just talking about processing power here. AI demands improved cybersecurity, enhanced auditing and reporting, more flexible governance practices, in addition to robust data pipelines, scalable storage solutions, and network architectures that can handle the computational intensity these applications require.

Even more challenging is confronting your technical debt and all those shortcuts, workarounds, and "temporary" solutions that have accumulated over years. Investing in legacy system you've been meaning to modernize.  Reimplementing those integrations held together with digital duct tape. They won't magically become AI-ready just because you want them to be.

I encourage you to have a frank discussion with your technical teams about what needs to be addressed before you can responsibly scale AI initiatives. Yes, it might slow down your timeline. Yes, it might require significant investment. But building AI on shaky technical foundations is like constructing a skyscraper on quicksand – impressive until it all comes crashing down.

Data Pipeline: The Backbone of AI Quality

Then, it’s time to talk about data, the fuel that powers every AI engine. I've seen countless organizations pointing to huge amounts of data as proof that they're AI-ready. But what we’re now recognising is that having lots of data isn't the same as having good data for AI applications.

Sit down with your data teams and ask the hard questions. How clean is your data really? Can you trust its accuracy and completeness? Is your data infrastructure appropriate? Do you have proper data governance frameworks in place, or are you flying blind with inconsistent definitions and quality standards across departments?

The reality is that AI systems are only as good as the data you feed them. Poor data quality doesn't just limit AI performance, it can actively introduce bias, generate unreliable insights, and create compliance nightmares. Before you get swept up in AI's possibilities, ensure you have the data governance structures, quality controls, and privacy protections that will support responsible AI deployment.

This isn't glamorous work, but it's absolutely essential. Think of it as laying the plumbing before you build the house – it’s not exciting, but critical for everything that comes after.

The Human Factor: Preparing for AI-Driven Change

Finally, and perhaps most importantly, let's address the human side of AI adoption. Your technology and data can be perfect, but if your people aren't prepared for the changes ahead, your AI initiatives will struggle.

Have you been transparent with your team about how AI might change their roles? Not just the positive "AI will augment your capabilities" messaging, but the real talk about which tasks might be automated, which jobs might evolve significantly, and what new skills will become essential? There seem to be growing gaps in discussions about the reality of AI’s impact on jobs.

We know that employees appreciate honesty over false reassurance. They want to understand how to position themselves for success in an AI-enabled organization. Of course, this means investing in reskilling programs, creating clear communication about AI's role in your company's future, and involving your workforce in shaping how these tools get implemented. But, this starts with being open about why not everyone will take this journey, and how do deal with the fall out.

Too often organizations focus change management on training people to use new AI tools. That’s necessary but not sufficient. Succes requires helping people navigate a more fundamental shift in how work gets done.

Your Back-to-School Moment

Just like those September resets of our youth, this is also an opportunity to review our own working practices and habits. In recent days I've found that I needed to give myself a good talking to. Look in the mirror and ask some hard questions about my own leadership approach, working practices, and attitude toward AI. It may also be a good time for you to do the same.

Are you genuinely curious about understanding AI's capabilities and limitations, or are you just following the latest trend? Have you been making decisions based on hype rather than evidence? Are you asking the right questions in meetings, or just nodding along when technical teams throw around buzzwords? Sometimes the biggest barrier to successful AI adoption isn't infrastructure or data, it's leadership failings when don’t do the work to truly understand what we’re implementing and how to ensure it delivers value.

Take the time now to honestly assess these foundational areas, both organizational and personal. Have the difficult conversations with your teams, but also with yourself. Address the infrastructure gaps, data quality issues, and human concerns before they become roadblocks.

The organizations that will thrive with AI aren't necessarily the ones that move fastest. They're the ones that build most thoughtfully. This September, give yourself and your organization the gift of honest preparation. Your future AI-enabled self will thank you for taking the time to get the fundamentals right.