OpenAI Files for Trillion-Dollar IPO
OpenAI confidentially filed its S-1 registration statement with the SEC on May 22, setting the stage for what could become the largest technology IPO in history. The company, currently valued at $852 billion following its March 2026 funding round, is targeting a public listing between September and November 2026 at a valuation that could exceed $1 trillion.
Goldman Sachs and Morgan Stanley are leading the deal. The financial picture, however, is sobering: OpenAI generated $13.1 billion in revenue for full-year 2025 but burned through approximately $22 billion, resulting in a net loss of roughly $9 billion. In Q1 2026, the company lost $1.22 for every dollar of revenue — a ratio that will face intense scrutiny from public market investors.
The filing arrives alongside OpenAI's launch of DeployCo, a $4 billion consulting subsidiary designed to help enterprises integrate AI into their operations. As Simon Willison argued this week, both OpenAI and Anthropic appear to have found genuine product-market fit with their coding and agent products, with April 2026 marking an inflection point where revenue from AI agent capabilities began materializing at scale.
Trump Delays AI Oversight Executive Order Amid Internal Disagreements
The Trump administration postponed signing an executive order on AI oversight just hours before its planned ceremony on May 21, reportedly over provisions the president found objectionable. The delay came despite weeks of preparation and invitations already sent to tech executives.
The administration's pivot toward AI regulation — a dramatic reversal from its earlier deregulatory stance — was driven by mounting concerns over Anthropic's Mythos model, which demonstrated an unprecedented ability to identify and exploit cybersecurity vulnerabilities. Anthropic limited Mythos's rollout through a cybersecurity initiative called Project Glasswing, but tensions escalated when the company refused to grant the Pentagon unrestricted access to the technology.
Meanwhile, an Axios investigation revealed that the administration had significantly weakened CISA, the nation's top cybersecurity agency, precisely as AI-powered hacking capabilities emerged — raising questions about whether the government has the institutional capacity to manage the risks it now acknowledges.
Vatican Releases First-Ever Papal Encyclical on AI
Pope Leo XIV released Magnifica Humanitas on May 25, a sweeping 42,300-word encyclical — the first papal document ever dedicated entirely to artificial intelligence. The document goes beyond familiar concerns about job displacement and misinformation to identify what Leo XIV sees as the deeper danger: that humans may begin to view themselves and others as products to be optimized.
The encyclical argues that AI should serve humanity “not by tempting us to escape limitation through optimization, but by supporting a life of openness and communion.” It addresses autonomous weapons, privacy violations, algorithmic bias, and the dignity of work, framing these issues within a broader theological vision of human flourishing.
Notably, Anthropic co-founder Chris Olah attended the encyclical's presentation and published remarks engaging with its arguments — signaling that the document is being taken seriously within the AI industry itself.
KPMG Deploys Claude to 276,000 Employees Worldwide
In one of the largest enterprise AI deployments to date, KPMG announced a global alliance with Anthropic to roll out Claude across its entire workforce of 276,000 employees in 138 countries. The deployment embeds Claude — including the new Cowork and Managed Agents features — directly into KPMG's Digital Gateway platform, the firm's core client delivery infrastructure on Microsoft Azure.
The initial focus targets Tax & Legal services and private equity clients, where KPMG says that building AI agents for tax regulation compliance has gone from weeks to minutes. Anthropic also named KPMG a preferred partner for private equity, and the two companies plan to build Claude-powered products for PE portfolio companies. Full platform implementation is expected by September 2026.
The deal underscores the accelerating enterprise adoption trend: Anthropic's run-rate revenue has surpassed $30 billion, with over 1,000 business customers now spending more than $1 million annually — a figure that doubled in less than two months.
EU Moves to Streamline AI Act Rules and Extend Deadlines
The European Council and Parliament reached a provisional agreement on May 7 to simplify the EU AI Act, extending compliance deadlines for high-risk AI systems and clarifying existing requirements. The amendments also introduce new rules specifically addressing AI-generated intimate content.
The move reflects a pragmatic recalibration: as the AI industry moves faster than regulators anticipated, the EU is choosing to give companies more runway rather than risk a compliance crisis. Meanwhile, in the United States, state-level regulation continues to advance independently, with Colorado and Connecticut both passing significant AI governance legislation in May.