Altman walks back AI ‘jobs apocalypse’ warning, says white-collar displacement under-delivered

TL;DR:

• OpenAI CEO Sam Altman told a Commonwealth Bank of Australia conference in Sydney he was “delighted to be wrong” that AI would eliminate entry-level white-collar jobs at the pace he had expected, and now believes a “jobs apocalypse” is unlikely.
• Altman said he and his executives had been “roughly right” on the technological predictions made at ChatGPT’s 2022 launch but “pretty wrong” on the social and economic implications, citing a personal experiment in which he reverted to handling Slack and email himself after AI auto-replies under-delivered.
• The softening lands amid OpenAI’s preparation for a confidential US IPO filing in the coming weeks, with the company reportedly aiming for a $1 trillion (£740 billion) valuation and raising at least $60 billion (£44 billion).

The comments are notable because they directly contradict the displacement narrative now being acted on by several UK and global lenders — HSBC, Standard Chartered and CBA among them — that have publicly attributed job cuts to AI, and they will be quoted in UK economic-policy debates on AI’s labour-market trajectory.

Context and Background

Altman’s softening on jobs is partial: he was careful to note “it still may”, reserving the option to revert to the alarm position. The structural argument he advances — that the “human part” of employment is harder to outsource than he expected — is more interesting than the headline. His specific example, AI-drafted Slack replies that he stopped using because “we really do care about our interactions with people”, maps directly onto the customer-experience differentiation case Octopus Legacy made today for its AI-assisted probate launch (covered separately by Resultsense).

The position contrasts with the explicit AI-displacement claims being made by major UK financial-services employers. HSBC’s CEO has told staff “don’t fight AI” as the bank cuts roles; Standard Chartered’s CEO apologised earlier this month for the “upset caused” by AI-displacement comments; Amazon has publicly attributed white-collar reductions to AI automation. The UK Office for Budget Responsibility estimates AI could materially affect 40% of the UK labour force over the next decade, with the majority of occupations complemented rather than substituted — a framing closer to Altman’s revised view than to the executive rhetoric.

For UK readers, the relevant question is whether the substantive picture matches Altman’s or his banking-CEO peers’. UK Office for National Statistics labour-market data through Q1 2026 shows some softening in entry-level professional-services roles but not the structural-displacement signal the alarm position would predict. The IMF, OECD and Bank of England have all published more measured estimates in the past year. Altman is now closer to that mainstream view than he was a year ago.

Looking Forward

The softening matters for UK policy because OpenAI’s positioning shapes the Overton window for AI-labour conversations. With Altman publicly conceding white-collar displacement is slower than feared, UK ministers and the Tony Blair Institute (which has anchored optimistic projections) gain rhetorical room to focus on complement-not-substitute framings. Expect the contradiction with UK banking executive statements to be raised in the next round of Treasury Select Committee oral evidence sessions on AI-and-jobs in the UK financial sector.