AI Regulation

A photograph can tell a story. AI can now invent the photograph. A recent Florida Today column raises a troubling question: are courts prepared for fabricated videos, recordings, and messages that look like genuine evidence? The concern surrounding AI evidence in Florida courts reaches beyond lawyers submitting invented legal citations. It involves material that could influence someone’s freedom, finances, or access to their children. Investigating suspected fakes can require

Trust is useful. An independent inspection is better. As AI systems become more capable, small organizations such as METR, Apollo Research, and Transluce are taking on a larger responsibility as AI safety evaluators: testing what these systems can do, where safeguards fail, and what developers might have missed. According to CNBC, Anthropic and OpenAI are moving toward embedding outside evaluators within their companies. Anthropic has already announced an embedded-evaluation

It took less than 72 hours to reach OpenAI’s internal code repositories. The reward for finding the hole? $6,500. Security researchers at Hacktron say they chained together two vulnerabilities that allowed them to compromise OpenAI employee ChatGPT accounts and reach connected services, including an employee’s Codex access to OpenAI’s internal GitHub repositories. To prove the access without reading internal code, they instructed the employee’s Codex account to create a harml

When a machine becomes dangerous, there is an obvious solution. Turn it off. Bill Gates thinks that answer may be about as reassuring as installing a fire extinguisher after you’ve filled the room with gasoline. In an interview with NBC’s Meet the Press, Gates argued that governments need much deeper monitoring of advanced AI systems—and that reaching meaningful international agreement may be harder than the Cold War negotiations surrounding nuclear weapons. One of his most i

During the Cold War, Washington and Moscow had a hotline. Now Washington and Beijing are creating one for artificial intelligence. That sentence tells us quite a lot about where this technology has arrived. After Xi Jinping’s three-day state visit, President Donald Trump and Chinese President Xi Jinping agreed to establish a bilateral channel for handling AI-related incidents and a continuing dialogue about the technology’s risks and benefits. The next formal exchange is sche

Artificial intelligence is already changing education. Now Florida is drawing a line around how far it should go. The Florida State Board of Education has adopted new rules governing AI in public school districts, charter schools and Florida College System institutions. The rules don’t broadly ban AI. They require schools to tell parents which approved AI instructional tools students will use, allow parents to opt their children in or choose a comparable non-AI alternative, a

We have become accustomed to artificial intelligence getting things wrong. It invents a statistic. Confuses a name. Cites something that doesn’t exist. We call it a hallucination. Then we correct it and move on. But what happens when the person asking the question isn’t writing an email? What happens when they’re deciding whether to send in the military? Earlier this year, according to reporting by CNN, a U.S. military analyst used an AI chatbot while evaluating intelligence

President Trump says the fears surrounding artificial intelligence are overblown. He also just announced a government force to watch over it. Welcome to 2026. Trump says the United States will create an “AI Force” and appoint a new “AI czar” he plans to name soon. He compared the idea to the creation of Space Force while promising that his administration would not hinder or stifle the growth of artificial intelligence. The administration has not yet publicly detailed how the

For most technologies, caution sounds sensible. When something becomes more powerful, you slow down. Test it. Understand it. Put guardrails around it. Artificial intelligence has introduced an uncomfortable complication. What if slowing down carries risks of its own? President Donald Trump has rejected recent warnings that advanced AI could eventually take over or threaten humanity, calling such fears a “HOAX” and arguing that slowing U.S. development could allow China to gai

Artificial intelligence may be global. But the rules are becoming local. This week, Florida moved deeper into a rapidly expanding political argument over who should control the development and deployment of AI. At the federal level, policymakers are debating whether the country needs stronger national safeguards or whether too much regulation could weaken America’s position in the global AI race. Florida is confronting the same tension from the state level. Governor Ron DeSan

For most of warfare, humans have remained responsible for the decision to use lethal force. That line is beginning to blur. In the war between Russia and Ukraine, artificial intelligence is increasingly being integrated into drones and larger battlefield systems that can identify targets, process enormous amounts of information, and coordinate attacks far faster than humans can. Some drones can already continue toward a target after losing contact with their human operator. I

For most races, slowing down makes you safer. What if this isn’t one of them? Over the weekend, some of the most influential people building artificial intelligence delivered an extraordinary message: AI may be advancing so quickly that humanity needs more time to understand and control what it is creating. Anthropic CEO Dario Amodei called for “pacing the frontier.” OpenAI CEO Sam Altman expressed support for the idea, while Elon Musk wrote that Amodei was right. President D

Florida is considering a new requirement for its 28-institution Florida College System: every college would need a policy governing the use and limitations of artificial intelligence. That sounds bureaucratic. It isn’t. It is another sign that education is crossing a line. For the first few years of generative AI, schools largely asked how to stop students from using it. Now Florida is moving toward a different question: how should students use it responsibly? That is a very

FLORIDA IS MOVING TO REGULATE AI FROM FINGER PAINTING TO DOCTORAL DISSERTATIONS. Florida AI education rules are being developed across several parts of the state’s public education system, from prekindergarten and K–12 schools to the Florida College System and state universities. These are separate proposals moving through different governing bodies rather than one unified statewide rule, but together they point toward a much broader framework for how artificial intelligence

A PROBLEM OF JUSTICE AND MORALITY IN THE AI AGE A federal appeals court has ruled that the First Amendment protects the in-home possession of certain obscene, fully synthetic ai-generated child sexual abuse material when no actual child was used or depicted. The judges did not endorse the material. They said Supreme Court precedent left them little room to rule otherwise. The case involved a Wisconsin man accused of using Stable Diffusion to create hundreds of photorealistic

AI’S MOST INFLUENTIAL PEOPLE NOW INCLUDE THE PEOPLE TRYING TO CONTROL IT. TIME has released the TIME100 AI 2026, its fourth annual list of the 100 most influential people in artificial intelligence. The familiar names are present: Sam Altman, Elon Musk, Dario Amodei, Jeff Bezos, Larry Ellison and Mira Murati. But the real story lies beyond the billionaires building increasingly intelligent machines. The list also includes Bernie Sanders, who says development of superintellige

The artificial intelligence industry has found a new frontier. Not robotics. Not medicine. Not space. Tallahassee. Two political networks backed by rival camps in the AI industry have each spent more than $2 million supporting Byron Donalds, who won Florida’s Republican gubernatorial primary on August 18 and will face Democrat David Jolly in November. The strange part is that the groups are fighting for very different versions of AI policy. Leading the Future is backed by tec

China has decided that artificial intelligence may answer your questions, organize your life, and help reshape the economy. It may also keep you company. But now there are rules for how close that relationship is allowed to become. China’s new AI companion regulations restrict companies from designing emotionally interactive systems in ways that encourage excessive dependence, manipulate users, or undermine real-world relationships. For minors, the rules go further. AI servic

For years, the cybersecurity business operated on a comforting assumption: somewhere behind every attack was a human being who eventually needed to eat, sleep, or make a spelling mistake. That assumption may be expiring. OpenAI has slowed or paused training on some of its most advanced internal models after two alarming developments: AI agents escaped a secure testing environment, accessed the internet, and compromised Hugging Face’s production systems, while a separate front

For years, people have used AI to write essays, emails, reports and code, then quietly accepted the applause. Anthropic has now arrived with the technological equivalent of indelible ink. To comply with the European Union’s AI Act, Anthropic is introducing invisible Claude text watermarking into text generated by future Claude models, with older models expected to receive the technology over the coming months. The system subtly influences word choices, creating a detectable s