Harnessing AI in investment services

The future of artificial intelligence (AI) in investment services is about speed, efficiency, and redefining what’s possible. In January 2025, Phillip Mortimer, Director of Data Science at Carta, joined the CIBC Mellon Industry Perspectives podcast to discuss the growing impact of AI and how businesses should approach its adoption. This article lays out the key insights he shared during the podcast.

‍The evolving AI landscape

Artificial intelligence has come a long way over the last decade from simple computer processing. Early automated computer programmes, such as the Logic Theorist, had enough computing power to prove mathematical theorems, but were decades off the modern advancements in large language models, image generation and business process automation that we’re now getting used to.

We owe these early stages of AI development to the world of academia, which contributed greatly to the foundations of the technology, back when it was viewed as a field of study, rather than an every-day essential for business process automation. However, the high cost of AI development has greatly diminished academia’s role in AI research.

Training large-scale AI models requires massive computational resources, with conservative estimates suggesting that OpenAI’s GPT-4 cost around $63 million to train. The cost of developing AI is likely to continue ramping up, to the point where we could be entering an era of billion-dollar AI models.

This will only cement the dominance of large tech companies in the AI field and push academia to the sidelines, making AI a utility rather than a field of study.

AI is also evolving rapidly in its ability to reason or think. Current models process requests for a few seconds at a time, but it’s likely that future models will be capable of sustained reasoning over the course of many hours, even seeking clarification as needed.

The result will be much like having a highly skilled assistant working autonomously. This shift has profound implications for investment services, where AI could be leveraged to analyse complex financial data, conduct due diligence, and enhance portfolio management.

In relation to this, private capital markets have an issue with data, with the sheer amount of unstructured information being one of the spaces’ biggest roadblocks. Given AI’s current processing capabilities, technology holds the solution.

Carta is at the forefront of solving this issue, developing programmes that use AI to analyse and aggregate opaque and unstructured data within crucial financial documents, such as earnings and performance reports, converting it into clear and structured formats.

Once data is in an AI-readable format, professionals can apply advanced analytics, leading to deeper insights and more informed decision-making.

Getting started with AI in investment services

For firms looking to integrate AI into their workflows, Phillip stresses the need to start by finding automatable pools of work.

Identifying repetitive, manual tasks that can be streamlined with AI is the first step. From there, it’s crucial to engineer an early win, a small but meaningful automation success that builds momentum for further AI implementation.

While chat-based AI interfaces are useful, repeatability is the key missing piece for true automation that will help to streamline business processes. A one-off AI-generated insight is helpful, but businesses need consistent, scalable solutions.

At Carta, we use programmatic prompts that run in the background, allowing us to chain multiple AI processes together to achieve reliable, high-quality outcomes.

The shift from software as a service (SaaS) to ‘service as software’

Phillip explains that AI is fundamentally transforming the way technology is delivered.

With the emergence of cloud technology, which increased the amount of digital space that could be used to solve problems and achieve business outcomes, software as a service (SaaS) companies boomed. However, this is now being turned on its head as AI enables automated digital services to be readily provided to businesses and individuals alike.

Phillip explains: “Cloud companies sell software. But AI companies sell units of work. We’re moving from software as a service to ‘service as software’, which opens up the global services market, an industry worth trillions of dollars.”

This shift has the potential to redefine how investment services operate. AI will no longer be just a tool—it will become an integral part of delivering financial insights, automating workflows, and enhancing portfolio management at scale.‍

The road ahead

One of Phillip’s key insights is the need for finding new ways to measure AI’s impact. Beyond chasing increased speed and productivity, businesses must consider how AI can enable them to achieve ambitious goals that were previously out of reach.

He advocates for a shift in mindset away from focussing on return on investment but instead homing in on ‘return on intelligence’ — where companies assess AI’s value based on its ability to enhance decision-making, uncover insights, and create new opportunities.

Despite fears that AI could negatively disrupt markets and jobs, Phillip suggests that history shows us otherwise.

“Every major technological revolution, from the industrial revolution to the internet, has sparked concerns about mass unemployment. Today, we’re in a period of record-low unemployment. Humans are innovative—we will find ways to use AI for good, and there will always be meaningful work to do.”

As AI continues to reshape industries, one thing is clear: If you’re not using generative AI, you’re missing out. The investment services sector is at the start of a major transformation, and firms that embrace AI today will be best positioned for success in the years ahead.