As this week’s #FILSAmsterdam discussions highlighted, AI is no longer just a trading tool – it’s fast becoming the infrastructure of modern capital markets. This isn’t simply the next phase of automation: the real opportunity lies in how AI can fundamentally reimagine the OTC landscape (https://www.fi-desk.com/data-and-diversification-key-to-solving-liquidity-woes/).

If the past two decades were about digitizing trading, this decade is about delivering the intelligence to understand where and how to trade. AI now sits at the center of this transformation – it has become the fascia of markets: the connective tissue linking data, models, and execution. Just as biological fascia integrates muscles, AI threads through trading systems – harmonizing fragmented sources, maintaining structural coherence, and transmitting data across the ecosystem. The result is an environment where information flows freely, strategies align dynamically, and markets move with greater agility and precision.

But if the challenge now isn’t adoption – it is becoming “architecture”.  As @Chris hollands highlighted in his LinkedIn post (https://www.linkedin.com/posts/chris-hollands-37386b2_fixedincome-fils-activity-7384974121128001538-qq2m/?utm_source=share&utm_medium=member_ios&rcm=ACoAAARqE5IBeeTUq6qzU4Eqccq_UO6-OMeR6Ao) the task ahead is to build markets that are not only smarter with greater technological input – but also markets that are sovereign and auditable – best able to support the future trading desk. Here are this week’s top five developments in AI and trading:

1. Platforms to Agents – Exchanges Become Intelligence Hubs

The London Stock Exchange Group (LSEG) and Microsoft have deepened their strategic partnership to help transform how traders can access and use data. Through the open-source Model Context Protocol (MCP), LSEG will make over 33 petabytes of data securely available within Microsoft 365 Copilot and Copilot Studio, enabling traders to ingest real-time, AI-ready market intelligence directly into existing workflows. Teams can now build and deploy custom AI agents that interact with LSEG’s data and analytics to drive faster decisions, automated insights, and higher productivity all within familiar Microsoft tools. This reduces the friction between data, models, and execution, positioning LSEG as an AI middleware platform where models connect directly to live, auditable market data ( https://www.lseg.com/en/media-centre/press-releases/2025/lseg-and-microsoft-transform-access-to-ai-ready-financial-data-in-customer-workflows).

2. Compute as Market Structure – The New AI Infrastructure Arms Race

AI trading’s next constraint isn’t data – it’s power. This week saw a record $40 billion acquisition of Aligned Data Centers by a consortium including BlackRock, Microsoft, Nvidia, and xAI, creating one of the world’s largest AI infrastructure partnerships (https://www.reuters.com/legal/transactional/blackrock-nvidia-buy-aligned-data-centers-40-billion-deal-2025-10-15/). The deal highlights just how much compute, cooling, and connectivity are now as fundamental to markets as clearing and custody once were.

In Europe, Nscale (UK) committed to deploying 200,000 NVIDIA GB300 GPUs in partnership with Microsoft – extending the compute frontier from Frankfurt to Dublin. Allianz Trade reports EMEA’s AI and data centre construction pipeline is up 43% year-on-year. As the cost of latency falls and compute capacity rises, infrastructure itself becomes a competitive differentiator – determining who trains, trades, and transacts fastest – its not just trading faster, its trading smarter that will count (https://www.allianz-trade.com/en_global/news-insights/economic-insights/big-beautiful-data-centers-ai-infrastructure-second-wind-construction-sector.html) .

3. Market Structure & Integration – From Fragmentation to Federation

AI is accelerating the push toward greater integration, scale, and unified infrastructure in Europe’s capital markets. Euronext CEO Stéphane Boujnah recently posted his support for Chancellor Friedrich Merz’s call for deeper, more attractive European markets and greater consolidation under single supervision by ESMA (https://www.linkedin.com/feed/update/urn:li:activity:7384879352783618048/). Fragmented regulation and disjointed structures currently limit liquidity and hinder the effective deployment of AI across borders. A unified framework would allow AI systems to operate on consistent, high-quality data, driving smarter liquidity management, more efficient execution, and deeper price discovery. Consolidation isn’t just an economic goal but a computational one – automated trading thrives on scale, and fragmented liquidity is its greatest constraint.

4. AI in Oversight – Supervision as a Continuous Function

Regulators are fundamentally shifting from rule-based oversight to a test-and-learn approach that mirrors aviation safety systems—prioritising resilience over precision. The European Commission’s new Apply AI Strategy envisions continuous, cross-sector supervision of AI (https://digital-strategy.ec.europa.eu/en/policies/apply-ai), while the Bank of England’s blueprint promotes experimentation through programmable, data-rich market infrastructures powered by AI, DLT, and quantum computing (https://acrobat.adobe.com/id/urn:aaid:sc:EU:e7a0d276-1aaf-446f-a84b-98e0d1d0f948).  The FCA’s Smart Data Accelerator, built with Raidiam, extends the regulatory sandbox to let firms safely test real-world data sharing and interoperability – enhancing transparency, price discovery, and liquidity in secondary markets (https://www.fca.org.uk/news/news-stories/fca-announces-partnership-accelerate-delivery-open-finance). Together these initiatives mark a move away from static rulebooks towards greater live supervision, where AI models, trading systems, and data flows are continuously tested, versioned, and validated in real time – creating a regulatory environment that evolves alongside the technologies it governs.

5. Risk, Fraud, and the Edges of Autonomy

As AI becomes more deeply embedded in trading, the rules of the market – who participates, how they interact, and what defines legitimate behaviour – are being rewritten. Trust is evolving from a matter of reputation or regulation into a fundamental layer of market infrastructure. Germany’s financial regulator, BaFin, recently helped dismantle more than 1,400 fraudulent “AI trading” websites in a major operation (https://www.techradar.com/pro/security/over-1-400-websites-taken-offline-in-german-polices-operation-heracles), while the UK FCA’s prosecution of so-called “finfluencers” underscores the growing challenge of distinguishing genuine from synthetic market activity (https://www.fca.org.uk/news/press-releases/first-court-appearance-three-finfluencers-charged-fca-led-global-crackdown-illegal-promotions). We are shifting from a world where trading models assist traders to one where they are becoming active market participants – governed by rules, oversight, and accountability. As algorithms and humans now operate within the same trading fabric, supervision, authentication, and transparency are redefining what it means to engage – and trust – modern financial markets.

Thank you for reading. As always, I’d love to hear your thoughts – what you agree with, where you disagree, and what you’d like to see more of next time.

Best wishes,
Rebecca