The way that AI is being used by the investment management sector is shifting from a focus on operational efficiency to allowing for more sophisticated investment-centric applications. According to a Citi report, this is being driven by AI’s ability to handle very large amounts of data, generate meaningful insights, and automate increasingly complex tasks.
It’s worthwhile to note that AI in its current state of development has really come a long way since the early 2000’s. Despite considerable improvements (several orders of magnitude), AI algorithms still continue to make serious mistakes that can significantly impact the overall reliability of complex investment decision-making processes.
However, AI will most likely continue to become more advanced, but for now, it does require a significant amount of human intervention and monitoring to ensure accuracy.
The report from Citi has several important key takeaways:
- Generative AI and agentic AI are fundamentally reshaping investment management, moving beyond simple efficiency gains to enable advanced research, analysis, and decision-making by processing unprecedented data volumes and automating complex workflows.
- The focus of AI adoption in investment management has evolved from primarily improving operational efficiency to actively contributing to alpha generation through sophisticated data analysis and autonomous execution of tasks.
- AI applications are expanding across investment, distribution, and operations, with new use cases emerging and a growing “wishlist” for future potential uses, including research assistants, predictive analytics, and automated decision-making support.
- AI and GenAI are pivotal in accelerating the electronification of financial markets, transforming trade execution, data handling, and streamlining voice-to-electronic communications, particularly in less electronified asset classes like fixed income and derivatives.
- Despite rapid advancements, significant challenges persist, including concerns about over-reliance and cognitive debt, confirmation bias, data privacy, security, and the need for robust regulatory frameworks and talent upskilling.
- Investment firms face strategic decisions regarding developing AI capabilities in-house versus partnering with fintechs – the “buy vs. build” dilemma. AI accelerates the consideration for modularization of investment processes and leveraging external expertise where efficient.
- Future developments in AI for investment management include applications in longer-term strategic signal generation, deploying agentic AI for time series forecasting and uncovering complex relationships through graph neural networks.
AI has also emerged as a hey cybersecurity investment priority for firms.
Nearly eight-in-ten organisations say their cyber budget will increase in the coming year, as businesses contend with an array of cyber risks
Investment in AI was the main budget priority (36%) over the next 12 months, ahead of cloud security (34%), network security (28%) as well as data protection (26%). This, according to an update from PwC.
Cyber skills deficits weigh in as well: a lack of knowledge in the “application of AI for cyber defence (50%) and lack of relevant skills (41%) were the top two challenges over the last 12 months in implementing AI for cyber defence.”
