Pension funds confront the question of who owns AI

Global pension funds are creating dedicated AI roles in senior parts of their organisations, in a sign that the technology’s enterprise and investment potential, alongside the governance risks it may carry, can no longer be managed through existing roles alone.

In two high-profile appointments this month, Canada’s OTPP nabbed the New York-based global head of engineering at Macquarie Asset Management, Feifei Wu, as its senior managing director of investment technology and applied intelligence. Meanwhile, Australia’s largest pension fund AustralianSuper lured Microsoft’s local chief technology officer, Sarah Carney, to become its first head of AI and automation.

Both roles are newly created with Wu set to work closely with OTPP’s investment team “to ensure technology enables key business outcomes” and establish appropriate governance. Carney, meanwhile, will identify areas within AustralianSuper from investments to member services that can be strengthened by AI.

It signals that allocators are pouring resources into finding the right leaders to oversee AI capabilities with the potential to transform all facets of fund operations including investment management, internal productivity, member services, cybersecurity, liability management and compliance.

With that operational complexity comes the question of where AI accountability should sit within the organisation.

Most commonly the chief technology and operating officer is the executive responsible, and this is the case at both OTPP and AustralianSuper. Australia’s largest fund is somewhat unique as it no longer has a chief operating officer role after a restructure last October, instead employing chief platforms officer Mike Backeberg who leads the fund’s technology, digital and data services strategy.

For CPP Investments, Canada’s largest fund, and a leader on AI integration, chief operating officer, technology and operations Jon Webster is responsible for AI oversight. In an interview with Top1000funds.com in January Webster warned peers to be wary of using frameworks “off-the-shelf” from other organisations without evaluating them in their own context.

While CPP Investments is not yet at the point where it will overhaul its organisational structure because of the impact of AI the value of people in the investment chain will no doubt shift with the future development of AI, Webster said.

In an insight paper released in 2024, CPP Investments said it has conducted “strategic talent planning” at every level of the organisation as a part of its adoption of AI tools, indicating that the fund is already examining how AI and automation might impact its future hiring activities.

Chief executives rarely hold direct responsibility for AI implementation in asset owner organisations, but some consider it a personal leadership project with the most prominent case being Norges Bank Investment Management CEO Nicolai Tangen.

Of asset allocators, NBIM has one of the most bullish views on AI’s enterprise potential – much of it driven by Tangen’s top-down directive that using AI is not an option, but a must. It has helped build a company-wide AI-enabled culture with more than 70 “AI ambassadors” scattered across global offices, who volunteered from different teams to spread AI knowledge and explain use cases.

With that said, there is little to no public evidence that any asset owner has tied AI or related metrics to their executive incentive plans.

Some funds, such as CalPERS, include operational effectiveness as a part of their remuneration incentives but AI’s implementation and adoption are not yet a common performance metric for fund executives.

This gap stands in contrast to the industry’s stated ambition around the strategic importance of the technology and indicates an evolution in its adoption and integration into key initiatives.

Climate is an instructive parallel to demonstrate that funds are willing and able to incorporate initiatives into their executive remuneration structures should they be of strategic importance. Both OMERS and PSP Investments in Canada, for example, have linked incentives for their senior management teams to the progress of climate targets.

Funds also need to be aware of the evolving regulatory frameworks and expectations around the responsible usage of AI.

This month, the Pension Regulator in the UK, which oversees workplace pensions, released its AI plan, stating that pension trustees not only need to have clear governance and accountability around their own internal AI usage, but will also be held responsible for ensuring external service providers have similarly robust AI guidelines in place.

Australia’s prudential regulator APRA also wrote to super funds, banks and insurers in April that “the speed at which [regulated] entities can identify and patch vulnerabilities [related to AI] needs to operate much faster”.

“We cannot be blind to the risks of such powerful technology – whether in our own hands or the hands of those with malign intent,” APRA said in the letter.

As AI becomes a tool for improved operational efficiencies and investment insights, asset owners will need to evolve the accountability around their internal processes and usage of the technology.

They are also demanding stronger AI governance from portfolio companies, often voting against directors and filing shareholder proposals to promote stronger transparency, and the credibility of their positions may depend on the robustness of their own frameworks.