London Tech Week 2026

Yesterday was my first time attending London Tech Week and my first thought was.. this could just be called “London AI Week”. No matter the keynote, panel, or conversation, everything kept coming back to the same topic: AI.
And not just AI products or LLM hype, but AI’s role in governments, regulation, sovereignty, customer interaction, ethics, and ultimately the future of work itself.

A few themes really stood out to me throughout the day.

Europe is not trying to replace the US in AI. It’s trying to coexist differently.

One of the most interesting discussions came from the keynote:
“The Emerging Case for Sovereign AI Development in Europe”, with George Osborne, Judith Dada and Matt Harris.

The conversation around “AI sovereignty” was fascinating because it wasn’t framed as “Europe vs the US” but instead, the message was more nuanced:
Europe likely won’t compete head-on with American AI giants on raw model dominance, but there is a growing desire to build sovereign European AI ecosystems that operate under European laws, governance, privacy standards, and societal expectations.

In other words: partnering with major US players while ensuring Europe maintains strategic control over how AI is deployed locally. That feels like a much more realistic than trying to fully replicate Silicon Valley.

AI is moving from being part of the customer journey… to owning it entirely.

Another keynote that stood out was: “The Future of Human Technology Interaction”, with Matt Clifford and Mati Staniszewski from ElevenLabs.

Today, AI is often integrated into parts of the customer support journey. Tomorrow, companies want AI to handle the entire journey.

We are moving from “AI-assisted support” to “AI-native interaction”.

And when you hear companies like ElevenLabs talk about voice AI becoming almost indistinguishable from human interactions, it becomes very clear that we are already dee in that shift.

The people building AI models should not be the only ones deciding what is “safe”.

One of the afternoon keynote: “Securing the Future of AI: The Roadmap to Safe and Ethical AI” was probably one of the most thoughtful discussions of the day. The panel included standout researchers and eperts on AI models and how they perform/are utilized in the real world.

One key point that resonated strongly with me: it probably should not be solely the responsibility of AI companies themselves to define what “safe” or “ethical” AI means, but external parties should be involved (as they’ll not be as biased). That conversation feels increasingly important as AI capabilities accelerate faster than regulation.

Another interesting concept discussed was the distinction between:

“aligned AI systems” → systems designed to operate according to human values, societal rules, and intended objectives
“narrow AI systems” → systems optimized for highly specific tasks rather than attempting general intelligence

It reinforced the idea that the future of AI may not simply be “bigger models”, but more specialized, aligned, and controlled systems. Not everyone needs the best, most powerful and most expensive AI.

Ironically, AI may make human skills even more valuable.

One of the most refreshing talks came from Jayney Howson at ServiceNow. Her argument was simple but powerful: AI itself isn’t the competitive advantage, but human skills will become the differentiator.

Because if everyone eventually has access to similar AI capabilities, then the real value shifts toward:

communication
creativity
judgment
adaptability
leadership
emotional intelligence
the ability to ask the right questions

As someone working closely with startups and hiring markets every day, that point especially resonated with me. The companies that win probably won’t just be the ones with access to AI but the ones that combine AI with exceptional human talent and decision-making.

Overall, the biggest takeaway from the day was that we are no longer discussing how and when AI will transform industries. The conversation has shifted toward: How do we govern it? How do we integrate it responsibly? Who controls it? And what remains uniquely human once AI becomes everywhere? Very curious to see where these conversations are one year from now.