- Pope Leo XIV’s encyclical “Magnifica Humanitas,” issued May 25, 2026, demands legal frameworks and independent oversight to prevent AI from becoming “an instrument of domination, exclusion and death.”
- A CBS News/YouGov survey conducted May 13-15, 2026, found most Americans believe AI will reduce available jobs, while a Change Research national survey found those expecting mostly negative AI impacts outnumbered optimists by nearly 2 to 1.
- Connecticut’s Senate Bill 5, advanced in late May 2026, requires disclosure for automated employment decisions and bans AI companion use by individuals under 18, the most detailed state-level AI safety framework in the US to date.
Pope Leo XIV’s first encyclical describes AI as one of humanity’s greatest current threats and demands legal frameworks and independent oversight to prevent it from becoming an instrument of domination. Published May 25, 2026, the document lands as American public opinion on AI hardens: job displacement fears are widespread, confidence in government regulation is low, and state legislatures are moving faster than Washington to fill the void.
Vatican’s Urgent Call on AI Ethics
The encyclical, titled “Magnifica Humanitas: On Safeguarding the Human Person in the Time of Artificial Intelligence,” was timed to mark the 135th anniversary of Pope Leo XIII‘s Rerum Novarum. That framing is deliberate. By anchoring the document to the Church’s foundational text on industrial labour, Leo XIV is positioning AI as this era’s defining social challenge in the same way factory conditions were a century ago, a moral imperative, not merely a policy question.
The encyclical denounces the “culture of power” driving AI development and calls for the “most rigorous” ethical constraints on the technology. It acknowledges AI’s potential benefits while warning that unchecked development risks deepening inequality, weakening human relationships and concentrating power in ways that would make it “an instrument of domination, exclusion and death.” On data ownership, the document is explicit: control “cannot be left solely in private hands but must be appropriately regulated.” The document calls for robust legal frameworks, independent oversight, informed users and a political system that does not abdicate its responsibility.
American Workers Fear Job Displacement
A CBS News/YouGov survey found most Americans believe AI will reduce the number of jobs available in the US, with younger respondents among those most concerned about their employment prospects.
The pessimism runs deeper than current conditions. A Change Research national survey, reported May 6, 2026, found that among respondents who expect AI to reshape their future, those anticipating mostly negative changes outnumbered those expecting positive ones by nearly 2 to 1. The dominant fear, by a significant margin, is job loss and economic displacement.
The 2026 Milken Institute-Harris Poll, published May 7, 2026, found that more than four in five Americans agree that “AI workforce readiness should transcend partisan politics,” with large majorities of Democrats, Republicans and Independents alike supporting government-backed workforce transition programmes. That breadth of cross-party consensus points to a political opening for federal action, one that has not yet been taken.
Mounting Concerns Over Privacy and Data Control
Connecticut’s Senate Bill 5 is the most detailed state-level response so far. Advanced in late May 2026, it establishes a broad AI safety and transparency framework built around two provisions that stand out. It requires disclosure when automated decision-making plays a prominent role in consequential employment decisions, so workers know when AI has influenced their career prospects. It also prohibits AI companion use by individuals under 18, directly addressing concerns about psychological and developmental harm to younger users.
Illinois has also acted, with a new AI transparency bill requiring annual third-party auditing for AI systems used in regulated contexts. These state measures share a common design logic: auditable compliance processes rather than voluntary policy disclosures. An annual third-party audit creates a paper trail that regulators can act on; a disclosure requirement without verification does not. For a closer look at how AI governance frameworks are being evaluated at the enterprise level, see our comparison of Zenity and IBM watsonx.governance.
Recent polling aligns with this legislative direction. Public concern about how personal data is collected, processed and used by AI systems is substantial, and state legislatures are beginning to match that concern with concrete obligations rather than guidance documents.
AI’s Broader Impact on Human Life and Dignity
“Magnifica Humanitas” goes well beyond economics. The encyclical frames unchecked AI development as a direct challenge to human dignity, warning explicitly that the technology risks becoming “an instrument of domination, exclusion and death” rather than a tool for human advancement. The document identifies three specific threats: deepening existing inequalities, weakening human relationships and concentrating power in a small number of hands. Its call for data ownership to be regulated rather than left to private actors extends that concern to individual autonomy over personal information.
Connecticut’s ban on AI companion use by those under 18 addresses one concrete version of these concerns, the psychological and developmental risks of AI systems designed to simulate emotional relationships with young users. Public polling reflects a similar unease, with widespread concern about how AI systems collect and use personal data pointing to questions that go beyond privacy compliance into questions of control and identity.
Regulatory Efforts and Future Directions
Federal movement has been slower. The White House released a National Policy Framework for Artificial Intelligence in March 2026, outlining nonbinding legislative recommendations for a unified federal approach. Nonbinding is the operative word: the framework sets direction but creates no enforceable obligations, and the path to comprehensive federal AI regulation remains unclear.
The gap between state action and federal inaction now defines the US regulatory landscape. Connecticut and Illinois are writing enforceable rules. Washington is writing recommendations. That asymmetry will create compliance complexity for companies operating across state lines, particularly as more states introduce their own frameworks, each with different thresholds, audit requirements and enforcement mechanisms. The practical question for businesses is not whether federal regulation is coming, but how many state-level regimes they will need to navigate before it arrives. The Vatican’s encyclical and the public opinion data behind it suggest the political pressure for stronger oversight is not easing. For more coverage of AI policy and regulation, visit our AI Policy & Regulation section.
