Top Stories
Vatican Publishes First Papal Encyclical on Artificial Intelligence
Pope Leo XIV today releases Magnifica Humanitas (“Magnificent Humanity”), the first papal encyclical devoted entirely to artificial intelligence. The document, signed on May 15 — exactly 135 years after Pope Leo XIII’s Rerum Novarum reshaped Catholic teaching on industrial-era workers’ rights — frames AI as a comparable civilizational inflection point demanding moral guardrails.
The encyclical will be presented at 11:30 a.m. at the Vatican’s Synod Hall, with an unusual co-presenter: Christopher Olah, co-founder of Anthropic. The pairing carries political charge. The Trump administration sanctioned Anthropic in February for resisting unrestricted military AI deployment, and the company is currently suing the government over alleged retaliation. Having the Vatican platform Anthropic’s co-founder alongside a document warning against military AI applications is a pointed choice.
Pope Leo has warned that AI must not “absolve humans of responsibility for their choices,” and the encyclical reportedly focuses on preserving human dignity, opposing autonomous weapons, and establishing moral boundaries for AI development. The deliberate parallel to Rerum Novarum signals that the Vatican sees this as a generational statement, not a policy paper — a moral framework meant to shape the debate for decades. Whether it moves secular regulators remains to be seen, but with 1.4 billion Catholics worldwide, the audience is already enormous.
Source: Vatican News, America Magazine, NPR
OpenAI Files Confidential IPO Papers, Eyes $1 Trillion Valuation
OpenAI filed a confidential S-1 with the SEC on Friday, setting up what could be the largest technology IPO ever. Goldman Sachs and Morgan Stanley are leading the deal, with a public listing targeted for Q4 2026 — possibly as early as September.
The company is aiming for a valuation between $852 billion and $1 trillion, up from its last private round. But the numbers underneath that headline are challenging: OpenAI generated nearly $6 billion in Q1 2026 revenue while remaining “deeply unprofitable,” losing roughly $1.22 for every dollar earned. Executives have privately expressed concern about financing future compute contracts after missing internal targets.
When the S-1 eventually goes public (roughly 15 days before the roadshow), it will answer questions investors have been asking for years: actual cash burn rate, unit economics for model serving, the breakdown between compute and talent spending, and how revenue splits across subscriptions, enterprise, API, and the new Codex products. The risk section alone should be illuminating — covering everything from existential safety concerns to ongoing litigation over psychological harms to national security implications.
Anthropic, for its part, is targeting an October 2026 IPO at a potential valuation above $900 billion. The AI industry is about to get a dose of public market scrutiny.
Source: Fortune, RoboRhythms
Anthropic Projects First-Ever Operating Profit on $10.9 Billion Q2 Revenue
Anthropic told investors it expects Q2 2026 revenue of $10.9 billion — a 130% jump from Q1’s $4.8 billion — and its first operating profit of $559 million. The projections, shared ahead of a funding round and reported by the Wall Street Journal on May 20, mark a significant milestone for a company that has consistently prioritized safety spending over margins.
The growth is being driven by enterprise adoption of Claude, with coding assistance and cybersecurity applications leading the charge. Claude Code alone crossed $1 billion in annualized revenue within six months of launch. But Anthropic itself cautioned that profitability may not hold for the full year, as planned compute spending increases could push it back into the red.
The numbers should be taken with a grain of salt — projections shared during fundraising rounds tend toward optimism. Still, if Anthropic actually posts a profit this quarter, it would be the first frontier AI lab to do so, fundamentally changing the narrative that building safe AI necessarily means burning cash indefinitely.
Source: CNBC, CryptoBriefing
Industry Moves
$67 Billion Utility Merger Signals AI’s Infrastructure Appetite
NextEra Energy announced on May 18 that it will acquire Dominion Energy for $66.8 billion in an all-stock deal, creating the world’s largest regulated electric utility. The primary strategic rationale: AI-driven power demand. The combined company will serve roughly 10 million customer accounts across Florida, Virginia, and the Carolinas.
This is the largest utility merger in US history, and it exists because of data centers. The deal signals that AI’s power consumption has moved from a talking point to the dominant force reshaping the energy sector. If you wondered whether the AI infrastructure build-out was real or hype, a $67 billion utility bet is a pretty clear answer.
Quick Hits
-
OpenAI releases Privacy Filter: An open-weight model for detecting and redacting PII in text, designed to run locally so sensitive data never leaves your machine. A welcome move from a company not historically associated with privacy-first thinking. OpenAI
-
Dataiku ships open-source PII proxy: Kiji Privacy Proxy sits between your apps and external AI APIs, detecting and masking 16+ categories of personally identifiable information before requests leave your network. Apache-licensed and self-hosted. Help Net Security
-
Hedy AI goes fully on-device: Meeting transcripts, summaries, notes, and coaching now process entirely on-device with no data sent to remote servers. The local-first AI movement continues to gain commercial traction. Yahoo Finance
-
OpenAI and Anthropic both targeting IPOs: With OpenAI eyeing September and Anthropic targeting October, the two leading frontier AI labs could go public within weeks of each other — setting up the most consequential tech IPO season since the dot-com era.
Worth Watching
The Vatican’s entry into the AI ethics debate is significant not for its policy specifics but for its framing. By placing Magnifica Humanitas in direct lineage with Rerum Novarum — the document that shaped Catholic social teaching on labor rights for over a century — Pope Leo XIV is making a bet that AI governance is a generational moral question, not a regulatory detail. The choice to present alongside Anthropic’s co-founder while the company fights the US government over military AI deployment adds an unmistakable political dimension. Meanwhile, OpenAI’s IPO filing will force unprecedented transparency from the most prominent AI company in the world. The S-1 will reveal whether the “grow now, profit later” model actually has a path to sustainability — and the answer will reverberate across every AI startup’s fundraising deck.