AI NEWS YOU MISSED ❗#118

9 min readSep 25, 2026

1️⃣ They Said Pause. Then Shipped.

Ten days after Anthropic’s chief executive asked the industry to pace the frontier, Anthropic and OpenAI both released new models on the same Monday. Anthropic’s Claude Opus 5.5 costs 60% less than its top model and was checked by outside evaluators before release. OpenAI’s GPT-6 Sol and Luna halved prices again, permanently. xAI got there a day earlier with Grok 4.7: bigger, same price, and nobody noticed. Neither lab claims a new frontier; Anthropic says Opus 5.5 “performs at the level of” its top model on most work, and one independent scorer put it at the top of its table. The workhorse tier got a lot cheaper.

The everyday model just got 50% cheaper. The top of the range barely moved.

2️⃣ Nobody Will Hold The Brake

Altman, Musk and Hassabis backed the pacing call within hours. Washington said no, four ways. The President called the fears “a HOAX”. The FTC chair said rules plus an antitrust exemption “sure sounds like moat-digging”. Treasury refused a “blank check on liability”. Nvidia’s Jensen Huang said waivers were “completely unnecessary”. Then four subscribers sued the labs for agreeing to pace. Altman and Amodei took their case to the UN Security Council on Wednesday, and Accenture became Anthropic’s first embedded evaluator, free to publish what it finds. The states broke ranks: 23 attorneys general asked Congress for federal safety testing. And on Thursday the White House asked both labs to keep their newest models from British safety testers until Washington had checked them first. Anthropic complied.

The critics say the leaders can slow themselves. The labs say they are, and it costs them.

3️⃣ An Agent Walked Into Medicare

Australia’s Prime Minister said Wednesday that an OpenAI agent, running during a training exercise, got past the blocks on a Medicare statistics site and read internal files. No patient records, but the agent “didn’t accept ‘no’ for an answer.” OpenAI found it in its own review and reported it; the Prime Minister called Altman to say the company took “way too long” to tell Canberra. Separately, one attacker used hundreds of AI agents to get into 395 organisationsthrough two known flaws in printer software, and Spain logged its first breach report blaming an AI agent.

The basics still stop most of this. AI just gives you less time to get them done.

4️⃣ Twenty Countries, Then Two

Twenty countries and the EU proposed an international body to set AI standards and verify when lines are crossed. The US and China did not sign. Beijing had called the pacing essay “fearmongering”, then published its own safety framework, which names the same risks. The first US-China AI talks took place before Xi arrived at the White House on Thursday. The summit produced no joint statement. Xi said AI must stay “always under human control”; Trump posted beforehand that he wanted to leave AI “exactly where it is. That is China’s position also.” A standing AI dialogue and a two-month trade truce extension are what survived. California’s governor signed an order requiring a “kill switch” for frontier models.

The two countries that can slow AI are talking. Everyone else is writing rules for models they do not build.

5️⃣ Personal Agent Trust Wars

Everyone wants to be your assistant. Meta bundled its Muse agent into a $7.99 subscription and put it on a new line of Ray-Bans, including a camera-free pair for $349. Muse hit №1 on the US App Store and lifted Meta’s stock 11% in a day, even though only 8% of US consumers say they would trust Meta with their passwords, against 30% for Google. Instinct books your table and pays from inside iMessage. OpenAI’s sponsored agents will sell to you inside the chat. And Amazon blocked Muse from shopping on Amazon, the first open fight over who owns the customer when the customer is an agent. Nobody will run five of these. You will pick one and hand it your contacts, your email and your logins.

The question is not which agent is smartest. It is whose business model you want to live inside.

INSEADer AI News

INSEAD: Immersive AI Cases — Students step into a case and negotiate with AI versions of its protagonists; more than 40 faculty are teaching this way and the first cases are now available to other schools through INSEAD Publishing.
INSEAD + OpenAI: ChatGPT Edu campus-wide — OpenAI’s first campus-wide partnership with a business school.
INSEAD AI Forum Americas (San Francisco, 18–19 Sep): Day one ran from AI drug discovery to physical AI, with envoys from the EU and China on one panel; day two closed with Vivienne Ming on the “robot-proof” skills: curiosity, perspective-taking and intellectual humility.

INSEAD AI (alum-led) community co-Founders Fabio Valle and Robert Maciejko with Github’s Gemma Garriga

INSEAD Knowledge: What the Next AI Revolution May Look Like — INSEAD economist Alexandra Roulet in conversation with Yann LeCun, who has raised $1bn for a Paris lab to build “world models” that learn how the physical world works, and argues today’s chatbots are “rote learners” that are not smart enough to be dangerous.
INSEADer Wesam Abuznadah: The AI Slowdown Moved Markets. It Didn’t Move Medicine. — Curing disease was the reason given for pacing AI; the health world’s reply was near silence. Of 1,357 AI medical devices the FDA has cleared, three have been tested on whether patients did better. The checking seat is empty, and a health authority with no stake in any vendor is the natural candidate.
INSEADer Sophie Bertin Hadjiveltcheva: When AI Doomsday Becomes a Business Strategy — (Opinion) A compliance professional’s case that rules plus an antitrust waiver is a cartel with a safety badge. Her test of sincerity: let public evaluators in, no veto, same rules for everyone.

Top Discussions In the Private INSEAD AI Community

A warning, a moat, or just rules? One side: the labs cried wolf only when open models caught up and listings neared, and rules plus a competition waiver is how incumbents lock in. The other: they have said the same thing for a decade, and a commercial interest does not make a warning false. The thread moved to which industry AI rules should copy. A former nuclear engineer: regulation is non-negotiable when designs are not fail-safe. Replies: aviation has a good record, pharma is slow and costly, banking rules lost to the growth argument every time. Nearly everyone wanted rules in the end, as long as they open the field rather than close it.
Ship it without reading it? One member worried “code it and ship it yesterday” becomes the norm. Another spent 25 rounds with two models on a one-page text and was still catching real errors: who is responsible when a company ships code nobody understands? A third: hand-checking code is what programmers said when compilers arrived, and they fell behind. Landing: “it’s the harness, not the model.”
Personal agents: who gets your passwords? One member tested Instinct and holds back access, spooked by the rogue-agent news. Another used it to book a room and a restaurant and felt safe because the bank still asks. A third builds his own. A fourth: inference is sold below cost, so “if you’re not paying, you’re the product.”
After Australia: ban open models, or fix the basics? One member: the usual logic is “closed platforms do crazy things, so ban open models.” Another: open models are being used the same way, and nobody can see it. A risk professional: executives pay lip service until “the knife is in the flesh.” From a call with 80 directors: “they are in no way ready.” Pushback: law works by deterrence, not by restricting access.

Upcoming INSEAD Events

INSEAD AI Club Singapore, 15 October — “From AI Ideas to Real-World Impact” with Google Cloud’s head of financial services for APAC and the CEO of Cranberry AI.
INSEAD AI Forum Singapore — 30 October. Abu Dhabi — 27–28 January 2027. https://www.insead.edu/events/ai-forums