AI in investment: Consensus positioning, fee pressure and spotlight on culture

AI is no longer sitting on the edges of investment management. It is already changing how information is gathered, how research is processed and how investment teams spend their time.

But the bigger question is not whether firms use AI. It is how they use it, where they draw the line, and whether efficiency gains come at the expense of judgement, challenge and accountability.

These were some of the central themes explored during a panel discussion moderated by Natalie Kenway, featuring Ashley Oerth, senior investment strategist at Invesco, Jenn Hui Tan, chief sustainability officer at Fidelity International, and Fraser Lundie, global head of fixed income at Aviva Investors.

For many firms, the first wave of AI adoption has focused on efficiency. Across research, sustainability analysis and portfolio monitoring, AI is helping investment teams process huge volumes of information far more quickly than before.

Tan described sustainability data as a clear example, noting that AI allows firms to absorb and incorporate large amounts of forward-looking and unstructured information “at a much quicker rate” and potentially change “the process of how you build conviction”.

In fixed income, Lundie said the transformation has been dramatic with tasks that previously consumed the majority of an analyst’s time now being completed in minutes.

“Previously anything up to 80% of your job was going through a 500-page offering memorandum and trying to piece together a backward-looking Excel spreadsheet,” he said. “That’s all now done in 15 minutes after a deal gets announced.”

The implications for productivity are obvious. But the panel repeatedly warned against viewing AI purely through the lens of cost reduction.

Oerth argued that AI should be seen not as a replacement for jobs, but as a tool to automate tasks and increase the productivity of skilled professionals. “It depends on how you yourself are you being a curious and critical thinking person,” he said.

Colleagues or AI tools for debate?

That distinction matters because investment management is fundamentally an industry built on trust, accountability and judgement. While AI models can synthesise information rapidly, they are also prone to bias, consensus thinking and overconfidence.

As Oerth noted, models are often designed to support a user’s line of thinking rather than challenge it. In practice, this creates a risk that investment professionals increasingly turn to AI tools instead of colleagues for debate, scrutiny and alternative viewpoints.

The panel warned that this could have significant implications for markets if firms become overly reliant on similar models and workflows.

Tan cautioned that AI may amplify existing systemic risks rather than create entirely new ones. “I think the potential of investors all using the same models, all thinking about risks in the same sort of way, all coming to the same alpha-driven outcomes, really, to my mind, increases the risks of overcrowding,” she said.

Lundie echoed this concern, warning that consensus positioning could become more severe and more rapid as AI adoption grows. “Not only will everyone be in the same trade, but they’ll get there quicker,” he said.

Greater importance on culture 

This raises important questions for fund managers about differentiation. If access to AI tools becomes commoditised, firms cannot rely on technology alone as a competitive advantage. Instead, the differentiator increasingly becomes how investment teams interpret, challenge and apply the information generated.

That places greater importance on culture, diversity of thought and internal challenge within investment teams.

Lundie explained that within Aviva Investors’ fixed income business, AI has actually increased the value of debate and contrarian thinking. By automating much of the manual groundwork, teams can spend more time interrogating assumptions and testing alternative perspectives.

“The meetings are way more interesting than they used to be,” he said. “We want to really get the more left-field views out there.”

The discussion also highlighted growing questions around accountability and governance. While firms are still experimenting with AI integration, the panel agreed there must be clear boundaries around where human oversight remains mandatory.

“For us, I think within investment management it’s fairly clear - it’s fiduciary duty that is your line,” Tan said. “AI can assist your decision making, but it cannot replace the need for human judgement and accountability.”

Lundie added that responsibility must ultimately sit with portfolio managers, firms and boards, regardless of how advanced the technology becomes.

Greater fee compression?

The economics of the industry were another unavoidable theme. AI-driven efficiencies are likely to intensify the fee pressure already facing active managers.

Tan described investment management as an industry already experiencing “structural price deflation”, adding that AI would likely “accelerate those trends towards greater fee compression”.

For firms, this creates a strategic choice. Should efficiency gains simply reduce costs, or should they be reinvested into research, talent and better client outcomes?

That question may define the next phase of competition in active management.

The panel ultimately concluded that AI alone is unlikely to outperform human investors, and they are ready to embrace the opportunities it brings. Instead, the firms best positioned for success will be those able to combine technological capability with scepticism, creativity and strong governance.

As Tan put it: “The differentiator will be judgement.”

In many ways, that may become even more valuable as AI becomes more widespread. Because in a world where everyone has access to the same tools, the real edge may no longer come from the algorithm itself - but from the people asking the right questions of it.