The Pope released the first ever papal document on AI today, Google's 24/7 AI agent arrives next week, and a new study finds AI-driven layoffs aren't generating returns.
Pope Leo XIV publishes "Magnifica Humanitas", the world's first papal AI encyclical
On May 25, Pope Leo XIV released his first encyclical letter, Magnifica Humanitas ("Magnificent Humanity"), dedicated to protecting human dignity in the age of AI. The launch event at the Vatican's Synod Hall included Christopher Olah, co-founder of Anthropic and head of AI interpretability research, alongside Catholic theologians and ethicists. The document is timed to the 135th anniversary of Rerum Novarum, the 1891 encyclical that shaped how the Church responded to industrialisation.
Why it matters: When the world's largest religious institution publishes its first formal moral framework for AI on the same day as a safety researcher from one of the leading AI labs, it signals that the ethics conversation has moved well beyond Silicon Valley. For students and professionals, this is the moment AI ethics stopped being an academic elective and became a cultural debate.
Gemini Spark: Google's 24/7 AI agent arrives next week for Ultra subscribers
Following its I/O 2026 debut, Google is rolling out Gemini Spark, a cloud-based AI agent that works on tasks in the background, around the clock, without you having to prompt it. It checks with you before taking major actions. Built on Gemini 3.5 and the Antigravity platform, Spark will be available next week to Google AI Ultra subscribers in the US. Separately, Google AI Mode has crossed 1 billion monthly users, with queries more than doubling every quarter.
Why it matters: An always-on AI agent is a meaningful shift from a tool you open to one that runs your errands while you sleep. The Ultra price point keeps it premium for now, but mass availability usually follows within months. Watch Spark closely, it is the template for what every phone assistant will become.
113,000 tech layoffs in 2026, and a Gartner study says the ROI isn't there yet
Over 113,000 tech workers have been laid off across 179 companies in the first five months of 2026, with Meta, Cisco, and Coinbase among the most recent. Nearly half of tracked layoffs are being explicitly attributed to AI automation. But a new study by Gartner, covered by Fortune, found a troubling disconnect: companies are cutting headcount regardless of whether the AI they deployed is actually generating returns. Even OpenAI's CEO acknowledged "some AI washing where people are blaming AI for layoffs they would otherwise do."
Why it matters: Jobs are being lost to the expectation of AI efficiency, not always to proven efficiency. Entry-level and generalist roles are taking the steepest hit. If you are in tech or adjacent fields, this is the case for building AI skills, not because AI will protect your job automatically, but because it will become the baseline requirement for proving your value.
US Take It Down Act is now law, platforms have 48 hours to remove intimate deepfakes
The Take It Down Act (TiDA) went into effect on May 19, 2026, creating the first nationwide US framework for intimate deepfakes. Covered platforms must now remove non-consensual sexually explicit AI-generated images within 48 hours of a valid takedown notice, and must have compliance procedures in place. Separately, the DEFIANCE Act, which passed the US Senate unanimously in January, allows victims to sue creators and distributors for up to $150,000 in damages, rising to $250,000 when linked to stalking or assault.
Why it matters: Two major laws now make intimate deepfakes both a platform-removal obligation and a civil liability. For students and educators, this is the shift from "AI can be misused" to "AI misuse has legal consequences." Other governments are watching this enforcement framework closely.
AI study tools are attracting VC again: Gizmo raises $22M, Oboe raises $16M
Two AI-native education startups closed significant rounds this month. Gizmo, a platform designed to make studying addictive through AI-personalised practice, raised $22 million in Series A funding. Oboe, which generates full personalised courses on any topic in seconds, raised $16 million. Institutional investors who largely retreated from consumer ed-tech in 2024–25 are now returning, drawn by the near-zero marginal cost of AI personalisation at scale.
Why it matters: AI is finally making personalised tutoring economics work, what once required a human tutor for every student now scales to millions. For students and teachers in India and globally, this wave of tools is arriving faster than most schools are prepared for.
India's AI build-out is impressive. The employment warning is getting louder.
India's IndiaAI Mission has provisioned 38,000+ GPUs at ₹65/hour (roughly one-third the global average cost) with 1,050 TPUs also available on the national platform AIKosh. But global equity research firm Bernstein issued a warning this week about a deepening employment crisis, specifically calling out AI's pressure on India's IT sector, with spillover effects into real estate and services. Both stories are running simultaneously, which is the tension worth tracking.
Why it matters: India is building serious AI infrastructure faster than most countries. But infrastructure investment and workforce transition need to happen at the same pace. The gap between the two is becoming a defining question for the decade.
The growing gap between AI investment and AI returns
Global AI spending is projected to hit $725 billion in 2026. At the same time, a Gartner study shows that many companies cutting jobs due to "AI automation" are not actually seeing the productivity gains they expected. Companies are making irreversible workforce decisions based on AI's potential rather than its proven performance, and the workers bearing the cost are those in entry-level and generalist roles who have the least ability to pivot quickly. Watch whether the pressure mounts on boards and executives to actually show returns, or whether the narrative of AI efficiency proves self-sustaining regardless of the data.
Brussels bans AI nudifier apps + EU AI Act's August content-marking deadline looms
Brussels moved to ban AI "nudifier" apps (tools that digitally undress people in photos) in the days leading up to today's papal encyclical. Separately, the EU AI Act's requirement for machine-readable content-marking on AI-generated material takes effect August 2, 2026, giving platforms just weeks to implement labelling infrastructure. Both moves signal that the EU is turning AI rules into enforceable obligations, not aspirational guidelines.