Louise Dudley Federated Hermes
Investors and companies must start to consider the broader social and systemic implications of the artificial intelligence roll-out in their long-term planning, said Federated Hermes portfolio manager Louise Dudley.
Speaking to Responsible Investor, Dudley said CEOs and boards need to proactively think about the ways in which their workforce and business models may need to evolve to remain fit-for-purpose.
“Currently, near-term pressures, such as higher oil prices and elevated interest rates are taking up the bandwidth of company boards rather than forward-looking planning on issues like AI which may be to their detriment,” said Dudley, who leads ESG and responsible investment research strategy within the global equities team.
“We do see a big spread in terms of the companies that have identified and are targeting explicit financial upsides and those talking about a handful of cherry-picked examples.”
Companies that start earlier are more likely to be in a position to take advantage of the benefits of AI, she said, citing the advantages gained by European companies that built processes on data governance and privacy ahead of the EU GDPR deployment.
Federated Hermes is particularly aware of the “potential systemic risks that could emanate from a very concentrated group of providers and suppliers”, Dudley said.
Investors have a lack of transparent information around pricing and usage stats needed to price AI companies, she said, while the firms also wield influence through government lobbying and have major impacts on the job market – such as driving demand for specialised engineering roles and potentially displacing entry-level jobs.
A useful guide for investors looking to get a handle on their exposure is to look at the adoption rate within different industries, Dudley explained.
In professional services this may come to relatively high levels, at around 70 to 80 percent adoption, she said, while things may change more slowly in other sectors.
At the moment, she believes the rhetoric from industries with high levels of AI adoption that they will maintain and reskill their workforce people is, for the most part, true – certainly in regards to the European employment model.
Dudley said investors can also look to frameworks like the Silicon Boundaries – co-developed by RI columnist Jakob Thomä and Oxford University’s Ben Caldecott – to guide their thinking on AI’s long-term societal impacts.
Looking ahead
Over the next 18 months, Dudley expects AI conduct risk – where AI systems take actions leading to unfair, illegal or non-compliant outcomes – to emerge as a material sustainability risk.
“If you’re working with a diverse supply chain, the risk of cyber-attacks and fraud massively increase, so it is important that AI models are fully vetted and checked, in addition to the authority and permissions that these models are granted,” she said.
A potential blind spot is in the value chain of large-caps, which may themselves have strong policies in place but lack oversight on their suppliers.
The Pennsylvania-based asset manager, which merged with Hermes Investment Management in 2018, has become an outspoken investor voice on the responsible use of AI and issued the EOS Digital Governance Principles as a reference on the topic.
The article has been updated to correct Louise Dudley’s title from responsible investment research head to portfolio manager.