🧭
You don't need a tech background to read this. Every term is explained.
World
Story 1 of 4
The Pope Wrote a 42,000-Word Letter About AI -- and He Is Worried
What happened
On May 25, Pope Leo XIV published a major document -- called an encyclical -- about artificial intelligence. He warned that AI is becoming too powerful too fast, that it threatens human dignity and democracy, and that governments and companies need to slow down and put rules in place.
Think of it this way
In 1891, a previous Pope named Leo XIII wrote a similar letter about the dangers of the Industrial Revolution -- factories, child labor, unsafe conditions. That letter helped shape laws protecting workers around the world. This new Pope chose the same name and is trying to do the same thing, but for AI.
Encyclical: One of the most important types of documents the Pope can publish. It goes to Catholic bishops worldwide and is considered major official Church teaching. With over a billion Catholics globally, this is not a small statement.
Why this matters to you
When an institution with over a billion followers calls AI the defining moral challenge of our time, it changes what politicians feel they are allowed -- and expected -- to do. This gives governments cover to regulate AI more aggressively. One of the scientists invited to speak at the Vatican event works at Anthropic, the company that makes Claude. That detail matters: the Church is talking to the people actually building these systems.
United States
Story 2 of 4
Illinois Just Passed the Toughest AI Safety Law in the Country
What happened
Illinois lawmakers passed a bill -- called Senate Bill 315 -- that would force the largest AI companies to hire independent outside auditors to check whether their AI systems are safe. The bill passed with zero votes against it. The governor says he will sign it.
Think of it this way
When a restaurant wants to open, a health inspector -- someone independent, not hired by the restaurant -- checks the kitchen. The restaurant cannot just say "trust us, it's clean." Illinois is trying to create the equivalent of a health inspection system for the most powerful AI systems. Right now, AI companies mostly just tell us their systems are safe. This law would require proof from an outside party.
Third-party audit: An independent check of a company's systems or claims, done by someone who does not work for the company and has no financial reason to give a good result. The word "independent" is the key part.
Whistleblower protection: Legal protection for employees who report dangerous or illegal activity at their company. Without this, employees who raise safety concerns risk losing their jobs. The Illinois bill would make it illegal to fire someone for reporting AI safety problems.
Why this matters to you
The companies building the most powerful AI systems currently get to decide for themselves whether those systems are safe. Illinois is trying to change that. If more states follow, this could become the de facto national standard -- even without Congress acting. It is also significant that both OpenAI and Anthropic supported the bill. When the companies being regulated support the regulation, it suggests they see a benefit in having clear rules.
United States
Story 3 of 4
Colorado Had a Strong AI Law. Lawmakers Quietly Replaced It With a Weaker One.
What happened
Colorado originally passed one of the strongest AI laws in the country. This month, the governor signed a replacement that is much weaker. The new law only requires companies to tell you when AI is being used to make important decisions about you. The original law required companies to actively prevent AI from discriminating against people. That requirement is now gone.
Think of it this way
Imagine a city passes a law requiring restaurants to prevent food poisoning -- by requiring safe food-handling practices, regular cleaning, and staff training. Then, under pressure from restaurant owners, the city replaces that law with one that only requires restaurants to put a sign on the door saying "AI was used in preparing your food." The sign tells you something. It does not protect you.
Algorithmic discrimination: When an AI system produces outcomes that are unfair to certain groups of people -- for example, rejecting loan applications, job candidates, or housing requests at higher rates for people of a particular race, gender, or age. This can happen even when no one programs the AI to discriminate, because the AI learns patterns from historical data that may already contain bias.
Why this matters to you
Colorado's retreat shows how quickly strong AI protections can be watered down once companies lobby against them. The same pressure will be applied in Illinois and every other state that tries to pass meaningful rules. Whether Illinois holds the line is one of the most important things to watch in AI governance right now.
Europe
Story 4 of 4
Europe's AI Rules Are Now Official -- but the Deadlines Got Pushed Back
What happened
Europe has the world's most comprehensive AI law, called the EU AI Act. After months of negotiation, European lawmakers agreed on a package of changes that delay some of the law's requirements by over a year. Companies now have until late 2027 -- instead of August 2026 -- to comply with the rules covering high-risk AI systems. The law itself did not get weaker. Companies just got more time.
Think of it this way
Europe built a very strict set of building codes for skyscrapers. The deadline for existing buildings to meet the new codes was originally this summer. Now it has been pushed to late 2027. The codes are just as strict. The buildings still have to comply. They just have more time to make the changes. For companies operating in Europe, the homework is the same -- the due date moved.
EU AI Act: A law passed by the European Union that puts rules on how AI can be used, especially in high-stakes areas like hiring, healthcare, law enforcement, and education. It applies to any company whose AI affects people in Europe -- including US companies.
High-risk AI systems: The EU's category for AI that makes decisions with serious consequences for people -- like whether you get a job, a loan, a visa, or medical treatment. These systems face the most rules under the EU AI Act.
Why this matters to you
The EU fine for breaking these rules can be up to 3% of a company's total global revenue. For a company like Google, that is billions of dollars. This means European rules affect the AI systems that Americans use too -- because companies build for global compliance, not jurisdiction by jurisdiction. When Europe sets a floor, it tends to become everyone's floor.
The Big Picture This Week
AI is developing faster than the rules meant to govern it. This week, we saw a Pope, a state legislature, and an entire continent all trying -- in their own ways -- to slow that gap down.
What is missing is anyone coordinating at a global level. Every actor is moving independently. The Vatican issues moral guidance. Illinois passes a state law. Europe extends its deadlines. The US federal government fights state laws in court while doing nothing itself. No one is in charge of making sure these pieces fit together.
That gap -- between the speed of AI and the speed of governance -- is the story. It gets more important every week.
Quick Reference: Terms Used This Week
- Encyclical
- A major official letter from the Pope, sent to Catholic bishops worldwide. One of the most authoritative forms of Church teaching.
- Third-party audit
- An independent safety check done by someone outside the company, with no financial stake in the result.
- Whistleblower
- An employee who reports illegal or dangerous activity at their company. Whistleblower protections make it illegal to punish them for reporting.
- Algorithmic discrimination
- When an AI system treats certain groups of people unfairly, even if no one intended it to.
- EU AI Act
- Europe's comprehensive AI law. Applies to any company whose AI affects people in Europe.
- High-risk AI
- The EU's category for AI used in consequential decisions -- hiring, healthcare, law enforcement, education.
- Trilogue
- A negotiation between three EU institutions: the Parliament, the Council (member states), and the Commission. All three must agree before a law changes.