The financial markets have never suffered from a lack of data. If anything, the challenge for modern traders and investment managers is quite the opposite: they are drowning in it. From real-time pricing and news feeds to unstructured earnings call transcripts and social media sentiment, the volume of information is immense. The critical differentiator in today’s market is no longer access to data, but the ability to cut through the noise to find actionable insights at speed.
A recent white paper commissioned by LSEG, Using AI for Rapid Access to Robust, Accurate and Usable Insights, outlines how financial institutions are addressing this challenge by using Artificial Intelligence (AI) as a fundamental tool for data discovery, interoperability, and workflow optimisation.
The Challenge of Data Overload and Fragmentation
The sheer volume of data has created an industry-wide problem of ‘noise’, further compounded by fragmentation; content is often accessed across multiple platforms and applications, leading to integration frictions, costly bespoke technology builds, and slow deployment.
Furthermore, a significant portion of valuable context exists in unstructured formats – PDFs, emails, and transcripts – which are notoriously difficult to ingest and analyse systematically. Documents like client contracts were previously stored separately from structured metadata, requiring months of manual review. Now however, AI can scan thousands of documents in minutes to extract key metadata, such as risk ratings and industry codes.
Rethinking Discovery with Natural Language
One of the most transformative impacts of AI explored in the paper is the shift in data discovery. Traditionally, retrieving specific datasets required knowledge of technical syntax or complex platform navigation. AI is replacing this with Natural Language Processing (NLP), allowing users to interact with data using plain language.
Bart Joris, Head of FX Sell-Side Trading Proposition at LSEG, notes that AI allows platforms to “look into the intent of the conversation or query to bring relevant data towards the users”. This capability effectively bridges the gap between ‘trader language’ and the technical code required to make an API call.
Rather than a user having to actively search for specific data points in separate apps, AI can infer context and push relevant insights directly to the user.
Interoperability and the ‘Chat-to-Trade’ Workflow
The white paper highlights that AI is not just about finding data; it is about acting on it within a seamless workflow. By integrating data into collaboration tools like Microsoft Teams, LSEG is facilitating workflows where collaboration and augmentation converge.
A prime example of this is the “Chat to Trade” function. In this scenario, a user can discuss terms with a counterparty in a chat window and initiate a regulated trade directly using a command, without ever leaving the conversation. This interoperability significantly reduces the time spent switching between disparate applications, boosting efficiency and reducing the risk of error.
Joris cites an example involving Tradefeedr, an FX Analytics platform. By using AI to process historical trade data and real-time market data, traders can benchmark their execution against the market. The AI interprets a natural language request for benchmarking data within a chat and retrieves the necessary structured data immediately.
Unlocking Unstructured Data
For investment managers, the ability to synthesise unstructured data is a game-changer. Advanced AI models can now process vast volumes of documents, such as earnings call transcripts or research reports, providing concise summaries in seconds.
Instead of reading a lengthy transcript, a user can ask a natural language question about how a company’s cost drivers have evolved over several quarters. The AI understands the intent, searches the unstructured content, and returns aggregated summaries. This does not just save time; it ensures that analysts can maintain a broader horizon view of the data without being overwhelmed by the reading load.
Hyper-Personalisation and User Experience
The future of the trader desktop is hyper-personalised. Akansha Goyal, Product Lead for LSEG Workspace AI and Personalisation, explains that the goal is to make the platform intuitive by analysing usage data and telemetry to push highly targeted content.
This goes beyond simple recommendations. It involves creating a single, immersive platform that proactively brings relevant information and interactive UI widgets to the user based on their inferred intent. Emily Prince, Head of Analytics and AI at LSEG, describes this as delivering “the right information to the right individuals at the right moment,” emphasising that the industry is only at the beginning of this journey toward true personalisation.
Trust, Governance, and the Human in the Loop
Despite the enthusiasm, the white paper underscores that successful AI deployment relies heavily on trust and governance. “Trusted data is the foundation of any successful AI use case,” states Prince. If the underlying data is limited or unverifiable, the AI’s output will be flawed.
There is also the complex issue of “agentic AI”—AI agents that can perform tasks autonomously. A practitioner at a Tier 1 UK sell-side institution notes that while the landscape is moving toward 24/7 automation, humans must remain “in the loop” rather than constantly “at the helm”.
A head of AI at a Tier 1 US fund manager raises valid concerns regarding agency: “Who makes decisions on behalf of whom? What happens if the action was not good and harms someone?”. Financial institutions must implement rigorous controls and independent data testing, similar to financial model stress testing, to ensure accuracy and prevent bias.
In conclusion, the integration of AI into financial markets represents a fundamental shift in how trading and investment decisions are made. By solving the problems of data overload and fragmentation, AI empowers professionals to work faster and smarter. Whether through summarising complex research, enabling natural language data discovery, or facilitating seamless trade execution within chat interfaces, AI is turning compliance data and noise into profit-generating assets.
However, as these technologies become embedded in decision-making, their long-term success will depend on robust data governance and the continued ability to combine machine speed with human judgement.
Subscribe to our newsletter
