AI risk debate: Steve Boese on what HR leaders can act on while consensus is missing

The global argument over AI’s risks is being fought at full volume, yet the decisions HR faces are more immediate. From readiness to governance, the H3 HR Advisors co-founder outlines where the function’s attention belongs.

September brought the argument over AI’s risks into the open from several directions at once. Early in the month, a researcher resigned from Anthropic, warning that the technology could lead to human extinction, and colleagues at the organisation publicly backed his concern.

The debate then moved to the United Nations (UN) General Assembly. Members of the UN’s International Scientific Panel on AI cautioned against apocalyptic rhetoric, and the next day US President Donald Trump rejected any global scheme to control the technology. A week later, he announced a voluntary accord with leading AI companies at the White House. Around the same time, Anthropic’s own IPO prospectus, as reported by Reuters, listed “catastrophic or existential risks to humanity” among its risk factors.

None of it settles what AI will do to jobs, which is the question most HR leaders are left holding.

Steve Boese, President and Co-Founder, H3 HR Advisors

Steve Boese, President and Co-Founder of H3 HR Advisors and Programme Chair of the HR Technology Conference, suggests HR leaders “treat these warnings about job losses as potential scenarios to examine rather than inevitable outcomes.”

In his view, the lack of agreement does not justify waiting. “The absence of consensus is not a reason to stand still, but it is a reason to proceed with care,” he tells HRM Asia. That care means testing AI against real work, building employee capability, strengthening governance and measuring what actually changes.

Reacting to the most dramatic forecasts carries its own risk. “Organisations that react only to the loudest predictions about disruption risk either moving recklessly or failing to build the capabilities they will eventually need,” Boese adds.

Boese has followed HR technology through several cycles, and parts of this one look familiar. Each wave, he notes, has arrived with high expectations and talk of rapid transformation, and each has underestimated how difficult organisational change really is.

The difference this time is scope. AI is moving past automating transactions into interpreting information, making recommendations, creating content and, increasingly, executing work. That gives it a much broader potential impact. Whether organisations capture that value, Boese says, depends on whether they can show they are ready to adapt and embrace genuinely new ways of working.

Then there is the matter of how AI is used to explain job cuts.

“AI can be a legitimate catalyst for redesigning work, but it can also become a convenient explanation for conventional cost reduction or decisions that were already underway,” Boese says.

Before accepting that framing, he suggests HR leaders look at which tasks have actually changed, what evidence supports the expected productivity gains, how the remaining work will be distributed, and where the technology’s value ends up.

“A credible AI transformation should involve a clear redesign of work, not simply a reduction in people followed by higher workloads for everyone who remains.”

READ MORE: HR technology’s 2025 story: Acceleration, accountability and the real test for AI

Who carries responsibility for that redesign is less clear-cut. Much of AI’s use in the workplace runs through HR, in hiring, performance management and workforce planning, while the technology itself usually sits with another function.

Boese sees accountability as shared across business leadership, technology, HR, legal and risk, but only up to a point. “It cannot become so distributed that no one is accountable for the workforce consequences,” he says.

For HR, that means being involved while systems and work are being designed, instead of being brought in afterwards to manage communications or employee concerns. Holding that role, he adds, requires stronger AI literacy, access to relevant data, clear governance and audit rights, and “the authority to challenge deployments that create unacceptable risks for employees or the organisations.”

Share this articles!

Latest Topics

More from HRM Asia

Promoted for results, left to figure out the rest

In the second of two pieces, Gwen Lockington, Founder and Principal of Lockington People and Advisory, explains why management capability remains a persistent gap, and what flatter structures and AI now ask of the managers who remain.

Subscribe to Our Newsletter

Stay updated with the latest HR insights and events,
delivered right to your inbox.

Sponsorship Opportunity

Get in touch to find out more about sponsorship and exhibition opportunities.