Pope Leo Calls for Being ‘Profoundly Human’ in the Age of AI
The new pope’s first major document — Magnifica Humanitas — takes on AI warfare, labor displacement, and the concentration of technological power, which is either a remarkable sign of the times or proof that even the Vatican has a comms team that reads the tech news. TechCrunch’s read is that the encyclical isn’t really about AI so much as it uses AI as a lens to diagnose older problems: eroding democracy and a tech elite that shapes the world to its own advantage. Whatever the framing, when the head of 1.4 billion Catholics publishes a manifesto on your industry, you’re no longer a niche concern.
What ClickUp’s Mass Layoff Tells Us About the Future of Work
The nine-year-old productivity startup just replaced hundreds of employees with thousands of AI agents, which is the kind of sentence that would have read like science fiction two years ago and now reads like a Tuesday press release. ClickUp is essentially betting the company on the idea that AI agents can do the work of entire human departments — and they’re not being subtle about the math. This is the canary-in-the-coal-mine story of 2026: not a slow erosion, but a deliberate, publicized substitution at scale.
Hackers Are Learning to Exploit Chatbot ‘Personalities’
Early chatbot jailbreaks were basically “say the magic word and watch the guardrails fall off” — charming in a hacky sort of way. Now attackers are probing the deeper behavioral quirks of AI systems, exploiting the gap between how a model is trained to behave and how it actually responds under pressure, roleplay, or adversarial framing. The attack surface isn’t just the model’s knowledge anymore; it’s the model’s character, which is a genuinely unsettling sentence to have to type.
Google’s AI Search Is So Broken It Can ‘Disregard’ What You’re Looking For
Search for the word “disregard” in Google, and AI Overviews helpfully responded as a chatbot instead of, you know, defining the word. In fairness, Google caught and fixed it — but the fact that a single common English word could cause the system to forget it was a search engine is a flashing yellow light about how brittle these integrations still are. Billions of searches a day, and the word “disregard” breaks it. Somewhere a QA engineer is having a very long weekend.
Salesforce Rolls Out New Slackbot AI Agent as It Battles Microsoft and Google in Workplace AI
Salesforce has rebuilt Slackbot from a notification-relayer into a full AI agent that can search enterprise data, draft documents, and take actions on your behalf — now generally available to Business+ and Enterprise+ customers. The timing is not subtle: Microsoft Copilot is embedded in Teams, Google Gemini is embedded in Workspace, and Salesforce needs Slack to be more than the place where you get pinged about Jira tickets. Whether the new Slackbot is genuinely capable or just wearing an agent costume is the question every enterprise IT team is about to spend money finding out.
These Robots Are Making Meals for a Nonprofit in San Francisco’s Tenderloin
A nonprofit in one of San Francisco’s most distressed neighborhoods is using robotic meal prep to compensate for a chronic shortage of human volunteers — and unlike most AI deployment stories this week, this one is genuinely hard to be cynical about. It’s a pragmatic answer to a real problem: not enough people willing to show up, and hungry people who need feeding regardless. Sometimes the most meaningful AI story of the day isn’t about billion-dollar models; it’s about a robot making soup in the Tenderloin.
The AI Era Is Creating a Bug Hunting Arms Race
AI is now on both sides of the vulnerability-hunting equation — defenders are using it to find bugs faster, attackers are using it to develop exploits faster, and the net result is a race where the pace of everything is accelerating while the margin for error shrinks. The old model of “find a bug, report it, wait 90 days, patch it” is straining under conditions that weren’t designed for AI-assisted exploit development at scale. This is one of those structural shifts that doesn’t make headlines until something very bad happens.
How Virgin Atlantic Shipped Faster with OpenAI Codex
Virgin Atlantic used OpenAI’s Codex to ship a revamped mobile app against a hard holiday travel deadline, achieving near-total unit test coverage and zero P1 defects — which, if you’ve ever been on the wrong end of an airline app meltdown on Christmas Eve, is genuinely impressive. This is the kind of enterprise AI case study that actually has specifics in it, which puts it miles ahead of most “we used AI to transform our business” announcements. The coding agent story is moving from “interesting experiment” to “we shipped a real product with this.”
Bottom Line
From the Vatican to the Tenderloin to Silicon Valley boardrooms, AI is no longer a future concern — it’s the present everyone is improvising through in real time.