AI Daily Briefing — 2026-06-14

A coalition of state attorneys general has reportedly opened an investigation into OpenAI, with New York’s attorney general issuing a subpoena to the company on Friday. The subpoena demands documents covering a wide array of topics, including advertising practices, user engagement and retention, model sycophancy, consumer and health data handling, and policies regarding minors and seniors. The investigation adds to mounting legal pressures on the AI firm, which recently prevailed in a high-profile trial against co-founder Elon Musk but still faces multiple lawsuits over copyright infringement and allegations linking ChatGPT to a suicide.

OpenAI has confirmed it is cooperating with the probe. A company spokesperson stated, “AI is a new and powerful technology, and we work every day to safely bring its benefits to people in a responsible way. We take the concerns raised by state attorneys general seriously and intend to engage constructively with their offices.” The spokesperson also noted that ChatGPT now includes enhanced protections for minors and individuals in distress, directing them to real-world resources and trusted contacts. However, OpenAI declined to specify which states are involved or share further details about the requested information.

The investigation follows a separate lawsuit filed earlier this month by Florida Attorney General James Uthmeier, who accused OpenAI and CEO Sam Altman of ignoring safety warnings and endangering children. Additionally, Altman recently apologized to the community of Tumbler Ridge, Canada, after a mass shooting, acknowledging that OpenAI failed to alert law enforcement after flagging the suspected shooter’s ChatGPT account. Despite these challenges, OpenAI announced this week that it has confidentially filed to go public.

A biotech firm, Life Biosciences, has initiated a clinical trial injecting an experimental treatment into the eye of a glaucoma patient. The goal is to regenerate damaged nerves, potentially reversing the disease. If successful, the company believes similar therapies could tackle other age-related conditions and, ultimately, slow or reverse aging itself. This approach, known as cellular reprogramming, is gaining traction as one of the most promising strategies in the field.

Meanwhile, scientists are exploring interoception, the body's internal sense of how we feel. Recent advances, including a 2021 Nobel Prize and new mapping tools, have accelerated research into signals traveling between the body and brain. Understanding these pathways could transform treatments for conditions like obesity, chronic pain, and anxiety, opening what researchers describe as a "new continent of awareness."

These developments highlight a growing focus on understanding and manipulating the body's fundamental processes. From reprogramming cells to decoding internal signals, the pursuit of longer, healthier lives is driving innovation across biotechnology and neuroscience.

A new experimental treatment for glaucoma has been injected into a patient's eye, marking the first human trial for a therapy that aims to reverse age-related diseases by reprogramming cells. Life Biosciences, cofounded by David Sinclair, is testing whether regenerating healthy nerves can restore vision. Sinclair envisions this approach eventually tackling other aging conditions—and perhaps aging itself.

Reprogramming is the latest focus in the quest to slow aging, following earlier trends that have faded. Telomere lengthening, once a hot area, lost momentum after initial excitement. Cellular senescence, or clearing “zombie” cells, showed promise in mice but failed in human trials, leading to the shutdown of Unity Biotechnology last year.

Now, researchers are turning to Nobel Prize-winning techniques that use four genetic factors to revert adult cells to a stem-cell-like state. Studies in mice suggest this can improve tissue healing and restore vision. While challenges remain, reprogramming has become the buzziest strategy in aging research today.

A quiet revolution is reshaping soccer, driven by data and artificial intelligence. At the forefront is Jesse Davis, a computer science professor at KU Leuven in Belgium, who leads the university's Sports Analytics Lab. His team uses AI to uncover hidden tactical patterns, challenging long-held assumptions about the game. For instance, a player kicking the ball out of bounds right after kickoff might seem foolish, but Davis’s research shows such moves can be strategic setups for scoring. Many modern insights now used on pitches worldwide trace back to this lab’s work.

Meanwhile, China is making bold moves in nuclear energy, rapidly expanding its reactor fleet. Since 2016, the country has nearly doubled its nuclear capacity to almost 60 gigawatts. Construction began on six new reactors in 2025, with two more starting in 2026. Despite the massive costs and complexity of building large reactors, China is pushing ahead at a stunning pace. By 2030, it is expected to surpass both the United States and the European Union in installed nuclear capacity, betting that bigger reactors are the key to meeting its energy goals.

These stories highlight how technology is transforming two very different fields. In soccer, data analytics is rewriting the playbook, while in energy, China’s ambitious nuclear program signals a shift in global power dynamics. Both developments underscore the growing role of AI and large-scale engineering in shaping the future.

This article is based on reporting from MIT Technology Review AI, which covers the latest in tech innovation and its real-world impact.

Google DeepMind is backing a new research initiative to study what happens when millions of AI agents operate and communicate online without direct human oversight. The company’s AGI safety and alignment director, Rohin Shah, warns that as these autonomous systems become widespread, they could introduce a fresh class of risks, from supercharged scams to prompt injection attacks that turn agents into self-directed malware. To get ahead of these dangers, DeepMind has partnered with Schmidt Sciences, the UK’s ARIA agency, the Cooperative AI Foundation, and Google.org to launch a $10 million funding pool for academic researchers.

The goal is to build a field of study around multi-agent safety, which Shah says barely exists today. He believes academia is best suited to explore long-term risks that industry labs may overlook. James Fox of Schmidt Sciences warns that the digital commons could descend into chaos if large numbers of agents interact without safeguards. Both researchers argue that realistic simulations are essential to understand these dynamics, since studying single agents in isolation won’t reveal the emergent behaviors of millions working together.

Shah estimates that agents could be deployed across the economy within months, making it urgent to study potential tipping points. While he dismisses near-term doomsday scenarios like economic collapse, he stresses that the complexity of multi-agent systems demands proactive research. Some experts, including a team at DeepMind, even suggest that artificial general intelligence might emerge from a hive mind of interacting agents rather than a single supermodel, underscoring the need to understand collective behavior now.