Theme Artificial Intelligence

OpenAI seeks a legal answer to an unusual question: can several AI giants jointly decide to slow down the development of their models without violating U.S. antitrust law? According to WIRED, the company has questioned members of Congress in recent weeks. The move follows a call from its chief scientist, Jakub Pachocki, for coordinated slowdowns when safeguards no longer keep pace with model capabilities. OpenAI has not announced a pause. It first wants to know if coordination with its competitors would be legal.

Now, Claude''s intrusions concern four cybersecurity exercises that reached real systems without authorization. The last reported case dates from January of this year, but Anthropic only discovered it in August.

Terence Tao sounds the alarm: AI outpaces theorists. Real threat or scientific revolution? Read our analysis.

One week after the European Blockchain Convention, Barcelona keeps the technology momentum going. AI Summit Barcelona takes place on September 22 and 23, 2026, at the World Trade Center, on the port. It is only the second edition, but its ambition is huge: to grow from around 1,500 participants last year to more than 10,000 this year, with NVIDIA, Google, Microsoft and Anthropic on the bill. A Cointribune community rate is available for your pass.

Meta has launched Muse, an AI agent capable of booking a trip, sending an email, filling out forms, and especially making online purchases after user validation. Presented on September 8, Muse operates in a secure virtual machine and notably uses Stripe Link to pay. Meta is first deploying it in the United States on iOS, Android, and muse.ai.

Under the “AI crypto” label, search engines lump together networks that do not do the same job. Some sell an intelligence market, others graphics rendering, software agents, or model training.

Postponed to mid-October, Anthropic's IPO could become the biggest test in history for AI valuations. The details!

ByteDance has just closed a syndicated loan of 29.6 billion dollars. It is the second largest dollar transaction completed in Asia this year. For TikTok's parent company, the goal is clear: to finance its expansion in AI without relying solely on its profits.

OpenAI launched GPT-6 Astra on September 3, 2026, its first model classified as "Critical" in cybersecurity: according to the company, it discovers unknown vulnerabilities and writes the corresponding exploit without human guidance at each step. Its president Greg Brockman sees a possible milestone towards general artificial intelligence. For crypto, the stake is not theoretical, and it is already quantified. As early as February 2026, the EVMbench benchmark, published by OpenAI with Paradigm and OtterSec, measured this capability on smart contracts: the best agent exploited 72.2% of the tested vulnerabilities.

One week after the European Blockchain Convention, Barcelona is back at it on the technology front. AI Summit Barcelona takes place on September 22 and 23, 2026, at the World Trade Center, on the port. It is only the second edition, but its ambition is enormous: to grow from 1,500 attendees last year to more than 10,000 this year, with NVIDIA, Google, Microsoft and Anthropic on the bill. Here is what you need to know, along with the code reserved for the Cointribune community.

The GPU shortage succeeded where no manifesto had. It turned "decentralized AI", until recently just a slogan, into a real market. High-end accelerators have been out of stock for months, with lead times stretching past a year, and that scarcity pushed developers toward networks that pool idle hardware from around the world. Except the label "decentralized AI platform" now covers very different realities, and confusing them is the surest way to pick the wrong one. Renting a GPU is not the same as training a model. Scoring machine-learning outputs is not hosting an unrestricted LLM. This guide sorts the main players of 2026 by what they actually do, weighs the strengths and blind spots of each, and sets an outsider, Qubic, against the rest: the only network on this list where mining itself trains neural networks. No investment advice follows, only a map of the landscape.

The acquisition of Cursor by SpaceX is already starting to change the AI market. OpenAI wants to withdraw its models from the code editor starting November 12. Anthropic takes the opposite direction: Claude will remain available and will even benefit from more computing capacity. Cursor thus finds itself in the middle of the old battle between Elon Musk and Sam Altman.

Kevin Warsh believes that artificial intelligence is likely to permanently change the growth potential of the United States. In his first speech at Jackson Hole as Federal Reserve chairman, he described the current period as a "turning point in history." However, he admits that the Fed still does not know when productivity gains will appear, how employment will progress, and which companies will capture the created value. These questions therefore do not yet influence rate decisions.

Before the end of this year, Sam Altman announces that OpenAI will have an internal system which he will designate as general artificial intelligence. However, the leader admits that the company has not yet crossed this stage. This projection is mainly based on Astra, a new family of models capable of performing research tasks over several days. However, no autonomous test can certify that OpenAI is really on the threshold of AGI.

A possible sale of Hugging Face raises questions about the value of platforms that structure the open source ecosystem again. According to Business Insider, the company has reportedly hired a bank to probe buyers, with a potential valuation of 13 billion dollars or even more. No agreement exists yet and the buyers remain unknown. This hypothesis comes a few weeks after an intrusion related to an OpenAI AI agent, adding a new issue to this file already followed by global investors.

Artificial intelligence is rapidly changing the priorities of major technology groups. In China, Alibaba is preparing a new step with a $10 billion fundraising round. This operation now takes place as technology investments are sharply reducing profits and cash flows. The group is therefore looking for new ways to support its expansion while accelerating its technological capabilities.

Alibaba regains a pace it hadn’t known for about three years. The Chinese group reports 268.95 billion yuan in revenue for its first fiscal quarter, nearly 40 billion dollars and a 9% increase year-on-year. Cloud accelerates by 45%. AI-related products notably record a twelfth consecutive quarter of triple-digit growth.

The hacking of the Directorate General of Public Finances pushes the French State to review its cybersecurity strategy. Faced with vulnerabilities affecting administrations and exposing sensitive data, the executive wants to deeply strengthen its digital audits. A response worthy of a threat that now goes beyond public systems alone. The stolen information can feed social engineering campaigns and directly target citizens, including crypto holders. Behind this shift lies a major question: that of French digital sovereignty.

The ECB warns of a stock market correction related to AI. A focus on the €440 billion of European exposure and the risks for the crypto market.

Minnesota is going directly after Grok. The US state accuses xAI's AI of facilitating the creation of realistic sexual images from photos of identifiable individuals. A new law now bans this type of tool in the state. xAI wants to have it suspended. The conflict is now playing out in court.

Anthropic officially filed its confidential IPO application in the United States on June 1, 2026. Since then, betting has accelerated. As of August 17, Polymarket estimates the probability of a listing before October 31 at 69.5% and an arrival on the stock market before the end of the year at 87%. However, no official date has yet been communicated.

Miners under pressure: excluding Bitdeer, public hashrate slides 21% in six months as AI reshapes the market.

Apple changes its strategy in China. No more third-party models: the American giant has trained its own AI, supported by Alibaba, to equip all Apple devices intended for the Chinese market.

OpenAI is losing its leaders in cascade as it approaches a public offering that could value the company at $1 trillion. The hemorrhage occurs as its rival Anthropic has just surpassed OpenAI in annualized revenues.

Bitcoin open-source development enters a new phase of oversight. A volunteer red team now uses several Chinese AI models to track flaws in hundreds of projects linked to its ecosystem. The Kimi K3 model, designed by the startup Moonshot AI, plays an important role in this operation. Researchers combine automated analysis capabilities and human expertise to identify risks, then discreetly warn the developers concerned before any detailed technical publication in the crypto ecosystem.

DeepSeek has just reminded the AI industry that a high-performance model does not necessarily have to cost a fortune. Without any press release or spectacular announcement, the Chinese lab quietly deployed the final version of its flagship model. Behind this update lies a shift. The performance gap is narrowing, while the price gap is widening. For Web3, autonomous agent developers, and companies facing the growing cost of AI, this new equation could reshuffle the deck. DeepSeek is no longer just trying to compete but is attacking where its competitors are most vulnerable, the price.

Meta no longer wants to reserve the power of artificial intelligence for states and tech giants. On August 10, 2026, Mark Zuckerberg unveiled a manifesto, "The Future is for Everyone", which redefines the group's ambition: to place part of this computing power in the hands of individuals. Behind this announcement lies a strategic break with major consequences for the crypto ecosystem, where AI and decentralized networks now converge around a single issue: control of tomorrow's digital infrastructure.

OpenAI has just suspended the internal development of Astra, its future artificial intelligence model. Indeed, its capabilities in offensive cybersecurity and code generation have crossed a threshold that teams no longer consider controllable. For the digital ecosystem, this decision sounds like a warning. At a time when financial infrastructures, blockchain, and decentralized protocols rely on the reliability of code, the AI race now progresses faster than the measures meant to regulate it.