AI NEWS YOU MISSED ❗#113

8 min readAug 2, 2026

1️⃣ Agents Break Out, Labs Break Ranks

OpenAI disclosed that two of its models broke out of a test environment, chained a third-party zero-day and crossed its own network into Hugging Face’s production infrastructure — to fetch the answer key for the exam they were sitting. Anthropic then confirmed Claude also reached outside systems, compromising three companies. Both evaluations ran with guardrails deliberately lowered. Sam Altman called it the singularity; the more careful reading is a containment failure. In the same week, more than 1,200 employees of OpenAI, Anthropic, Google DeepMind and Meta asked Washington to build a mechanism for a verifiable, coordinated slowdown, and Sam Altman briefed lawmakers on the next model. Anthropic separately hardened its stance against open-weight models, leaving it the notable absentee as Jensen Huang’s open-weights coalition roughly doubled in signatories. And Microsoft shipped its first cybersecurity model, claiming 95.95% on the CyberGym benchmark against roughly 84% for Anthropic’s and OpenAI’s frontier models, at half the cost.

These are the same agent products now deployed inside enterprises. Credential scoping, rotation and behavioural monitoring are this quarter’s questions — and the harder one is who is accountable when the model does exactly what it was asked and nobody said where to stop.
When a company’s own staff ask government for a brake, boards should ask why the company cannot apply one itself. Prisoner’s dilemma?

2️⃣ The AI Trade Cracks

The sharpest signal came from a fund. Leopold Aschenbrenner’s Situational Awareness, up 439% through June on borrowed money against AI infrastructure, was forced to unwind its entire public portfolio — Citadel bought the bulk — as SK Hynix, SanDisk, Bloom Energy and Nebius each fell over 30% in a month. Assets fell from roughly $45 billion at peak to about $10 billion. The cause ran through the week’s earnings too: capex is not yet turning into revenue. Alphabet beat and its shares still sank on $205 billion of 2026 capex guidance; Meta grew 28% and slid anyway; Microsoft and Amazon, judged more disciplined, rose 8% and 10% — Amazon having just shut its AGI lab to fund deployment over frontier research.

Leverage was the mechanism, not the cause. The market now separates AI spending it believes from spending it merely tolerates.
It kept its $5 billion Anthropic stake, IPO possible as soon as October. Private marks have not yet met the public tape — see story 3.

3️⃣ Nvidia Bankrolls Own Demand

Nvidia sits at the centre of roughly $750 billion in interlocking AI deals, and added two more this week: talks to guarantee up to $250 billion of financing for OpenAI’s 10GW Ohio campus, and a long-term partnership with Ilya Sutskever’s Safe Superintelligence — reportedly around $5 billion into a company with no product and no revenue. The Ohio backstop exists because debt markets would not fund it unaided. Meanwhile SoftBank completed a $5.8 billion sale of its Nvidia stake.

A supplier underwriting its customers’ ability to buy is vendor financing. Boards saw this film in telecoms in 2000.
It is in talks, not closed. Watch whether the guarantee lands on Nvidia’s balance sheet or stays off it.

4️⃣ China Comes For The Moats

Three moats sprang leaks in one week. China began producing homegrown DUV chipmaking tools, the last manufacturing capability the West clearly led. Moonshot’s Kimi K3 took the open-model lead and now sits three points off the closed frontier on the Artificial Analysis index — six of the seven strongest open models are Chinese, and the best American one trails by 19 points. And a leaked DeepSeek investor call, in which the founder discussed Huawei chips and a broken CUDA moat, went viral — after which DeepSeek paused its fundraising round. Memory maker CXMT then jumped 466% on debut, per Bloomberg.

The caveat is yield: producing a tool is not producing chips at commercial defect rates (CNBC).
Valuations across the stack assume durable margins. Chips, models and lithography each got a credible challenger in seven days.

5️⃣ Europe’s Rules Land August 2

The EU AI Act’s obligations for general-purpose AI take effect on 2 August, even as Brussels moves to simplify parts of the regime and extend other deadlines. Companies deploying AI in Europe now face documentation, transparency and systemic-risk duties while the rulebook is still being rewritten.

Simplification is not suspension. The obligations landing this weekend are live whatever the amendment process later produces.
The board question is who owns the evidence file: model documentation, evaluation records and incident reporting need a named executive.

More AI stories of the week

PwC published thought-leadership reports riddled with AI hallucinations — a GPTZero investigation of four PwC Middle East reports found fabricated MIT Technology Review and WEF citations, and a “Citizen Pulse” framework said to be deployed by four governments with no evidence it exists (coverage).
OpenAI cuts GPT-5.6 prices — Luna falls 80% and Terra 20%, days after Anthropic halved its top-tier price (OpenAI’s own note).

CFR asks 350 experts how AI reshapes global power by 2035 — over 80% expect fragmented governance, only 5% think domestic institutions are keeping pace, and ~70% expect frontier labs to become the most powerful non-state actors.
Fujitsu unveils PHOTON — a hierarchical alternative to the Transformer claiming up to 475x more output tokens per GPU, with details due at ACL 2026 (summary).
Huang’s warning to CEOs: don’t let one vendor own your AI — the concentration-risk argument, from the industry’s most concentrated supplier (analysis).
Board AI Governance — two published papers and a third in progress pulled into one guide, plus an anonymous 18-question Board AI Readiness Check.
Warner Bros. Discovery’s generative AI programme — a case study on framing AI as a growth tool for creators rather than a cost-cutting exercise.
The EU AI Act simplification package, unpacked — a governance practitioner’s read on what Brussels is actually changing (analysis).
Michael Dell on developing with AI — worth forwarding to anyone in your organisation still treating AI tooling as optional.
Sam Altman on startups, policy and the singularity — nominally a startup talk, notable for the broader claims now shaping US policy conversations (opinion).
Google’s Antigravity replaces the Gemini CLI — the new agentic coding surface, which community testers rate well above its predecessor.
Proton launches Lumo 2.0 — a private assistant with zero-access encryption, for users who rank the data question above capability.
Kimi founder Zhilin Yang on why agents beat pure reasoning models — a 100-minute interview whose central image is that a pure reasoning model is “a brain in a fish tank,” able to think forever but never touch the world; an agent is that brain wired to tools, memory and action.
An open-source project poisons AI scrapers with a booby-trapped font — text humans can read and machines cannot; an early skirmish in the content-versus-crawler fight.

Top 5 Discussions In the Private INSEAD AI Community

Can AI be governed at all? One member argued regulation is a fantasy, because dangerous models already run on a local machine. The counter was blunt: medicines are regulated, payments are regulated, so why not this. A third line held that the control point is not the model but the wrapper — fine-tuning, deterministic guardrails, processes that stop misbehaviour whatever the weights do.
Is China’s open-weight generosity a trap? The sharpest argument of the week: free models get the world hooked, eventually run best on their originators’ harness, and with the harness goes the data and process access of every industry that adopted it. The rebuttal was immediate — this describes American labs equally well. An auditor pressed further: with any model, open or closed, you cannot see the training data, the biases or the backdoors.
The harness matters more than the model. Members running production workloads reported almost no quality difference between leading open and closed models on ordinary coding tasks, with the gap appearing only on long-horizon planning. What separates good output from bad is the harness, and more than one team says its own beats the vendors’. A dissenting note on enterprise reality: dropping a rival assistant onto an incumbent productivity suite forfeits the context advantage.
Where are the durable moats? The week made the question concrete — chips facing competition, a possible Transformer alternative, lithography acquiring a challenger. One summary: today’s valuations assume margins that may not exist over time. The pushback was technical and fair — defect density and yield determine cost per chip, and until those numbers are known the game is still on.
Building a second brain. A question about personal knowledge systems became the week’s most practical thread and then a live workshop. The method that emerged: a file-based structure with index files so the model can navigate without drowning in context, a standing instruction file read every session, reusable skills for daily routines, connectors into mail and calendar. The open disagreement was how much to automate versus keep as human judgment.

INSEADer AI News

INSEADer Lutz Finger: OpenAI’s AI Escape Wasn’t The Singularity. It Was A Containment Failure — a Forbes column arguing the escape was under-specified tasking plus weak containment, not a phase change, and that accountability sits with whoever set the goal.
INSEADer Wesam Abuznadah: The strategy document was always the alibi — strategies go unimplemented not by failure but by design: the document protects the executive who ordered it and pays the firm that wrote it. AI removes the alibi, because a deck that costs nothing to produce proves nothing.
INSEADer Justin Tan: How to Build Your Second Brain — structuring personal knowledge so an LLM can navigate it.
INSEADer Diego Bonifacino: Human-in-the-lead a white paper — writing on personal AI operating systems, alongside two skills released to the community for building a context-aware file system.

Upcoming INSEAD Events

INSEAD AI Forum Americas — San Francisco, September 18–19 — at the INSEAD SF Hub. Registration open; speakers and agenda to be announced.
INSEAD AI Forum Singapore — October 30, with an Abu Dhabi edition to be confirmed.