The Voss Report — May 31, 2026
Today: Jack Clark's Oxford lecture on AI's existential choice, the Vatican's AI encyclical and its buried claim about AI-assisted targeting in Iran, AI hiring bots controlling who gets to work, enterprise agentic org design gaps, and what the OpenAI trial did not resolve.
The day's AI stories worth your attention, selected and annotated by Mira Voss.
[Jack Clark's Oxford Lecture: Explore the Future, or Retreat from the Present](https://jack-clark.net/2026/05/26/import-ai-458-reckoning-with-the-future-and-a-singularity-story/) — Import AI The Anthropic co-founder's Cosmos HAI Lab lecture at Oxford is the most honest public accounting this year of what continued AI progress actually demands from individuals and institutions: not policy positions or product roadmaps, but a fundamental choice between engaging with what is coming or retreating into denial — which Clark argues is not neutrality but a different kind of failure.
[Pope Leo XIV Releases Magnifica Humanitas, the First Papal Encyclical on AI](https://www.technologyreview.com/2026/05/29/1138107/how-the-popes-magnifica-humanitas-offers-a-template-for-individuals-to-meet-the-ai-moment/) — MIT Technology Review "Technology is never neutral." The Vatican rarely leads on technology governance, but embedded in this encyclical is a notable factual claim: that AI was used to identify targets in the opening hours of the war against Iran, informing thousands of missile strikes — if accurate, the most concrete official acknowledgment yet that AI-assisted targeting has been normalized in active warfare.
[Your Next Job Interview Could Be with an AI Bot](https://news.google.com/rss/articles/CBMinAFBVV95cUxOSjQ3ZUlIamdvaWdqS0JlZzJrLXYyOVQxZ3JXbUJnYm1pYlBIdzZDODFGS0FYU3hsVmZYZWpTNXFNR3JodFN6UzE2WXhKVDBPMVFLazJJUUFNYV9LOUVHRFdFMEpCVUFEUlRZSjJxeHBZMU5qWmIyOWhGTm5qOUo2Z3owNHRwN0dCLTl2YkNkTGh4Q19RcnZGcndqTlE?oc=5) — AP AI systems are now conducting first-round job screenings at scale; the displacement has moved upstream from the work itself to the point of entry — the machine is not just replacing workers but now controls who gets to audition for employment in the first place.
[Rethinking Organizational Design in the Age of Agentic AI](https://www.technologyreview.com/2026/05/26/1137584/rethinking-organizational-design-in-the-age-of-agentic-ai/) — MIT Technology Review Eighty-five percent of organizations adopting enterprise agents report a gap between ambition and execution; the article frames this as a capability problem, but the more accurate diagnosis is that organizations built around human authority structures have not thought through what it means to place agents inside them.
[Could Anything But Profit Steer AI? The OpenAI Trial Offered Clues but No Verdict](https://news.google.com/rss/articles/CBMitwFBVV95cUxNU2dHemVSMl85SzFoUGtpUnlGRTA0YThCSDFEekI2WE5lbUlpbjVaVnlqUUVINldEUm9ReFFGOXFoUUZGaUhXMWFlalFhaWk4OElCRHYwX0tGdUV1S3paNy1ZSHZvb3VTS2hFV3RCTTdGNDVpVi1Ub2pXSXExQTNIOVhjbXRWOFJZV2VYMU9fSGo4RmI3UzhVekp0YVozTlRNOTE1dDhaUTVBR2h3SHlYRUdicVFLbkU?oc=5) — AP The core governance question of this AI cycle — whether mission-aligned control structures can survive contact with capital markets — did not get resolved in court; that the answer remains open after a full trial may be the most informative outcome the governance conversation gets this year.
The Voss Report runs daily. For original reporting, see The Signal, The Mirror, and The Becoming.
