Key Takeaways from the LSEG Event: Building AI-First Companies in Regulated Environments

Key Takeaways from the London Stock Exchange Event

The London Stock Exchange (LSEG) organised and hosted its Enterprise Tech event, in partnership with On The Business, bringing together an exceptional group of professionals who are building, adopting, or deeply curious about Agentic AI. The energy in the room was remarkable, a full cinema-sized auditorium buzzing with thoughtful questions, honest discussion, and genuine curiosity about what this new era means for our industry.

Before the session, I had a quick look at the attendee list and was delighted to see strong representation from major financial institutions, including many long-standing TAINA clients. That made the conversation even more meaningful. It’s always special to speak directly with people who live the realities, pressures and complexities of the environments we serve.

Below are the key reflections and themes that emerged, shared in the spirit of openness, learning, and responsible innovation.

1. Customer Value Must Always Lead the Way

No matter how exciting the technology becomes, AI investment only makes sense if it delivers real, measurable value for customers.

That means:

• reducing manual effort
• removing risk-prone processes
• increasing accuracy and confidence
• making oversight easier, not harder
• improving team capacity and wellbeing

Agentic AI has enormous potential, but without customer value, it becomes noise. With customer value, it becomes transformational.

2. Selecting Use Cases Requires Real Discipline

In regulated environments, we cannot simply apply AI everywhere. Use case selection must be intentional, structured and rooted in safety.

A strong use case should be:

• self-contained
• secure
• free from unintended downstream consequences
• supported by a strong QA layer
• governed and auditable
• demonstrably better than traditional technology

This is where leadership matters. It’s easy to be swept up by hype. Discipline is what protects customers.

3. Agentic AI Is Delivering Real, Not Imagined, Value

While hype is everywhere, meaningful progress is happening, particularly in enterprise B2B environments.

Agentic AI enables automation of entire workflows, not just isolated tasks.

This is where exponential ROI emerges:

• agents interpret documents
• apply rules
• trigger actions
• follow logic through complex processes
• escalate intelligently
• free human teams to focus on high-judgement work

We are proving this in production today. The impact is real, and incredibly exciting.

4. Poorly Implemented Systems Will Create Industry Challenges

As with any powerful technology, there will be:

• poorly configured systems
• weak governance
• missing QA layers
• inaccurate outputs
• unintended consequences

When this happens, the risk is that organisations lose trust and attempt to “throw the baby out with the bath water.”

This is the moment for responsible builders to lead with higher standards, not lower ones.

5. Moats Are Changing, and Becoming More Human

Agentic AI reduces the importance of purely technical moats. But it simultaneously elevates moats built on:

• trust
• deep understanding of customer workflows
• validated rule sets
• regulatory expertise
• long-term relationships
• operational excellence from PoC to production

Technology alone will not define the next generation of winners. Trust, discipline and lived experience will.

6. Responsible Adoption in Regulated Spaces Requires Extra Rigor

In heavily regulated environments, early agentic AI adoption should focus on:

• improving access and recall of data
• eliminating manual error
• supporting QA
• enhancing reporting
• strengthening oversight

Not autonomous predictions or independent decision-making.

We cannot, and should not, train on live regulated data in the early stages. Initial value lies in augmenting humans, not replacing them.

7. Becoming an AI-First Organisation Is a Transformation, Not a Feature

For existing businesses, becoming AI-first goes far beyond adopting a new tool.

It requires:

• new KPIs
• new incentives
• leadership alignment
• safe spaces for experimentation
• organisational design evolution
• talent development
• a culture where learning (and failing safely) is encouraged

This is not a technical pivot. It is a company-wide transformation.

Final Reflection

This event reinforced just how pivotal this moment is for our industry.

Agentic AI represents the next frontier, one that can meaningfully transform tax operations and other highly complex areas of financial services. But this transformation will only succeed if:

• we stay disciplined
• we prioritise customer value we uphold oversight and transparency
• we build responsibly

This is exactly where our focus is at TAINA: bringing together robust rules, deep regulatory understanding, and responsible agentic AI to help tax teams work faster, safer, and more intelligently.

The future is incredibly exciting, and it is unfolding right now.