On May 25, 2026, Anthropic co-founder Chris Olah stood inside the Vatican and told an audience of religious leaders, policymakers, and technologists that AI will likely displace human labor at very large scale. Two days later, OpenAI — as explored in the intelligence factory race between AI labs — CEO Sam Altman said the opposite: he was delighted to be wrong about earlier projections that AI would wipe out entire categories of jobs.

This is not a philosophical disagreement. It is a strategic divergence with business model implications worth hundreds of billions of dollars.

Two Companies, Two Narratives

Anthropic has leaned into a safety-first narrative since its founding. By aligning publicly with the Vatican and Pope Leo XIV — who called AI a potential instrument of domination, exclusion and death — Anthropic is doing something no other frontier lab has attempted: building institutional trust with non-tech power centers.

OpenAI, meanwhile, is playing the optimism card. Altman’s message to Commonwealth Bank of Australia CEO Matt Comyn was calibrated for enterprise buyers and governments who want assurance, not warnings. If you are selling $200/month subscriptions and enterprise API contracts, existential risk rhetoric is bad for business.

The Strategic Logic Behind Each Position

Anthropic’s Bet: Regulation as a Moat

Anthropic benefits from regulation. The company has positioned itself as the responsible lab — the one governments should trust, the one the Vatican invites to speak. If AI regulation tightens globally, Anthropic’s safety-first brand becomes a competitive advantage. Stricter compliance requirements raise barriers to entry and favor incumbents who already invest heavily in alignment research.

The timing is not coincidental. Anthropic just hired OpenAI co-founder Andrej Karpathy to lead a new pre-training research team, and has been systematically recruiting top talent from OpenAI, Google, xAI, Microsoft, and Apple throughout 2026. The safety narrative attracts mission-driven researchers who might otherwise go to DeepMind or FAIR.

OpenAI’s Bet: Growth Through Optimism

OpenAI needs enterprises to accelerate adoption, not slow it down. The company is pursuing aggressive commercial expansion — multi-pronged agreements spanning GPU procurement, model deployment, and cloud infrastructure — as explored in the economics of AI compute infrastructure — with major tech companies. Fear of job displacement is the single biggest drag on enterprise AI adoption timelines.

Altman’s pivot from doomer to optimist also reflects a practical reality: OpenAI’s revenue depends on volume. The more companies that deploy AI agents across their workforce, the more API calls OpenAI processes. A jobs apocalypse narrative would freeze exactly the budgets OpenAI needs flowing.

The Talent War Underneath

While the philosophical debate grabs headlines, the real battle is playing out in recruiting. Anthropic’s hire of Karpathy — who started this week on the pre-training team under Nick Joseph — is one of the most significant talent moves in AI history. Karpathy co-founded OpenAI, led AI at Tesla, and built a massive following through his educational work at Eureka Labs.

This is not just one hire. Anthropic has landed at least ten high-profile recruits from OpenAI, Google, xAI, and Microsoft in 2026 alone. The company is reportedly poised to surpass OpenAI’s private market valuation.

What This Means for Business Leaders

• Vendor selection now carries ideological weight. Choosing Anthropic signals safety-consciousness; choosing OpenAI signals growth orientation.
• Regulation is coming regardless. The Vatican’s involvement signals that AI governance is becoming a global institutional priority, not just a Silicon Valley debate.
• The talent market is the real scoreboard. Whichever lab attracts the best researchers will build the best models. Right now, Anthropic is winning that race.
• Both narratives serve business interests. Neither company is being purely altruistic — each is positioning for the regulatory and commercial environment that best suits its strategy.

The Bottom Line

The Anthropic-OpenAI split on AI jobs is not about who is right. It is about who controls the narrative as governments decide how to regulate the most powerful technology since the internet. Anthropic is betting that being the cautious voice in the room builds trust. OpenAI is betting that optimism drives adoption. Both strategies are rational. Only one can be right about what comes next.

Read the full deep dive on Business Engineer →

Free tools: Map of AI | Try Business Engineering | Defensibility Audit

Subscribe free → businessengineer.ai