Masteria, Centre de formation IA certifié Qualiopi
Certifié QualiopiFinançable OPCOFrance · Suisse · Belgique
Français
Édition du Tuesday 22 September

Veille IAMore than twenty countries call for stronger human oversight of AI, without the United States, China or France, while Washington and Beijing open a warning channel between them ahead of the Trump-Xi summit. Google launches its Googlebook at $899 to sell Gemini, Meta's Muse agent races to the top of the charts before being blocked by Amazon and flagged for a security flaw, and British Columbia sues OpenAI after a mass shooting.

Par l'équipe éditoriale Masteria, sous la direction de Mathias Nizan · Publiée le · this morning

More than twenty countries call for stronger human oversight of AI without the United States, China or France, while Washington and Beijing open a warning channel between them ahead of the Trump-Xi summit. Google launches its Googlebook at $899 to sell Gemini, Meta's Muse agent races to the top of the charts before being blocked by Amazon, and British Columbia sues OpenAI after a mass shooting.

Trois zones suivies avec le même soin : l'Europe et la France, l'international, la Chine et l'Asie. Régulation, sorties de modèles, déploiements documentés, publications de recherche. Écrite par Masteria, organisme de formation et de conseil en IA, pour des dirigeants, des chefs de projet, des développeurs et des juristes.

13 actualités17 sources7 min de lecture
En bref
Au sommaire
Top story · Europe and France · International · China and Asia · Research and papers · In brief
Sources du jour
Le Monde · Journal du Net · ZDNet · TechCrunch · The Verge · Ars Technica · Bloomberg · Wired · South China Morning Post · arXiv
Analyse
Signée par l'équipe éditoriale, publiée avec chaque édition
Cadence
Publiée les jours ouvrés, vers 8h30
Corpus
344 actualités traitées depuis le 6 August 2026
Top story
Top story

More than twenty countries call for stronger human oversight of AI, without the sector's three great powers

The leaders of some twenty countries, including Germany and Canada, along with European Commission President Ursula von der Leyen, have signed a joint declaration calling for humans to stay in control, Le Monde reports. The text reflects a concern now shared at the highest levels of government: AI development could outpace our ability to keep it in check. The most telling absence is that of the missing signatories. The United States, China and France, three leading powers in the field, did not add their names. The dividing line is sharp between those betting on coordinated safeguards and those keeping their hands free, each for different reasons. A declaration without the players building the most advanced models works above all as a political signal.

SourceLe Monde
Today's stories

Les 12 autres actualités du 22 September

The 22 September edition covers 13 stories from 17 sources: 3 on Europe and France, 4 international, 2 on China and Asia, 1 research paper and 2 briefs.

Chaque actualité porte sa source. Le texte complet et l'analyse sont dans l'édition du jour.

Ouvrir l'édition complète

Research and papers

1 actualité

Aussi dans cette édition : 2 brèves.

Today's read

Amazon blocked Meta's Muse agent, Walmart quietly pulled its ChatGPT integration and retailers are rushing to build their own agents, while 31% of shoppers already prepare a purchase with AI: the moment of advice leaves the brand's own site for the model's answer, and the skill that gains value for a French marketing team is becoming the exact source that ChatGPT, Claude and Gemini cite, by fixing product data and measuring its share of citations rather than its rank in Google alone.

Amazon blocked Meta's Muse agent and Walmart quietly pulled its ChatGPT integration. Both moves say the same thing, and it speaks directly to French marketing teams: the moment of recommendation, when a customer decides what to buy, leaves the brand's own site for the model's answer. The Journal du Net puts it plainly: the brand keeps its content but loses the monopoly on the staging. It becomes a source that ChatGPT, Claude or Gemini cite or ignore, like any other. The numbers show where the shift stands. According to a survey relayed by ZDNet, 31% of online shoppers already use generative AI to prepare a purchase, while the traffic sent by autonomous agents rarely exceeds 1%. The gap draws the roadmap. Advice runs through AI long before the agent clicks in your place. Visibility in the answer is won today; automated buying will come next, and its plumbing is not yet laid, as shown by the doors that Amazon, Meta and Walmart are closing. The skill that gains value in a French organisation then changes in nature. Take your ten best-selling products. Ask ChatGPT, Claude and Gemini the question a customer would ask before buying, and read the answer. Look at whether your brand is cited, whether the price and specifications are accurate, which source the model opens. Where the information is wrong or missing, the cause almost always lies in poorly structured product data and unverifiable claims on your own pages. Fix the data at the source, not the campaign. Measure your share of citations in the models' answers, as you track your position in Google. The shelf has moved into the sentence the machine generates. The brand absent from that sentence no longer exists at the instant the customer chooses. Yesterday's search optimisation aimed at a results page; the one coming aims at a generated answer, and that answer is won through the accuracy of what you publish about yourself. The Masteria editorial team.

Amazon blocked Meta's Muse agent and Walmart quietly pulled its ChatGPT integration. Both moves say the same thing, and it speaks directly to French marketing teams: the moment of recommendation, when a customer decides what to buy, leaves the brand's own site for the model's answer. The Journal du Net puts it plainly: the brand keeps its content but loses the monopoly on the staging. It becomes a source that ChatGPT, Claude or Gemini cite or ignore, like any other.

The numbers show where the shift stands. According to a survey relayed by ZDNet, 31% of online shoppers already use generative AI to prepare a purchase, while the traffic sent by autonomous agents rarely exceeds 1%. The gap draws the roadmap. Advice runs through AI long before the agent clicks in your place. Visibility in the answer is won today; automated buying will come next, and its plumbing is not yet laid, as shown by the doors that Amazon, Meta and Walmart are closing.

Mathias Nizan, fondateur de Masteria
The Masteria editorial team
Under the direction of Mathias Nizan
Lire l'analyse complète
All editions

Previous editions

Chaque édition reste en ligne avec ses sources et son analyse. Elles sont classées par leur titre du jour.

Septembre 202614 éditions
Charles III gathers AI bosses in Scotland and warns of \"existential dangers\", while a Pew survey across 37 countries shows a worldwide fear of job destruction. Brussels adopts the EU KIDS Act, Mistral moves into Firefox, Nvidia, Google and Anthropic join forces to find 100 gigawatts on the power grid, and Huawei pulls its next training chip forward by nine months to close the gap with Nvidia.14 actualités · 8 min
OpenAI admits six model misbehaviours and publishes a framework to report them, the same day Google opens the connected home to agents and Anthropic folds its interfaces into a single Claude that decides for itself how much work to take on. The agent moves from reading to acting. Meanwhile Brussels adopts the EU KIDS Act to keep minors off social media, Mistral arrives in Firefox, and Huawei pulls its next AI chip nine months forward to cut China's dependence on Nvidia.14 actualités · 7 min
Silicon Valley refuses to brake, Washington splits14 actualités · 8 min
AI leaders' call to slow down splits the world13 actualités · 7 min
AI leaders call for slowing the race, the White House refuses: Dario Amodei argues for \"pacing the frontier\", Sam Altman, Elon Musk and Demis Hassabis agree, Donald Trump replies that he will not hand his lead to China, and Asian markets fall. Anthropic is eyeing the Nasdaq for a listing that could exceed $86 billion while OpenAI backs away from its own, SoftBank borrows $11.9 billion to fund OpenAI and Christine Lagarde warns about the financial cost of AI. In China, Z.ai returns to raise $5 billion in Hong Kong and the intelligence chief warns about the risks of AI.14 actualités · 8 min
AI safety moves from the activist fringe to mainstream debate: Sam Altman says he is ready to slow frontier AI development and is testing whether such a brake would be legal, a former Anthropic researcher warns that \"AI could kill us all,\" OpenAI adds a worried researcher to its board, and Anthropic publishes 161 pages of Claude misuse, from Yemeni missiles to fake dating profiles. In Europe, the CNIL sets rules for electronic invoicing and the EU's cybersecurity agency trials Anthropic's most offensive model. In China, DeepSeek reignites the price war with its V4.1 Flash model and chipmaker Enflame jumps 188% on its stock market debut. Microsoft plans 38 gigawatts of data centres and Jensen Huang promises Nvidia 70% growth.13 actualités · 8 min
Apple's first Ternus keynote and the OECD's student warning13 actualités · 7 min
OpenAI claims a Navier-Stokes proof, then a credit fight erupts14 actualités · 8 min
Mistral AI raises €3 billion at a €21 billion valuation in a round led by Samsung; the labor-law scholar Jean-Emmanuel Ray reminds employers that the AI Act requires them to inventory their high-risk AI systems and inform staff by the end of 2027, while BPCE stands up its AI Factory and Bercy renegotiates the French state's IT contracts; China targets 9,800 eflops of compute and Huawei launches its first stacked-circuit chip; Revolut completes the first AI-initiated payment in France14 actualités · 8 min
Moody's finds that Chinese companies get more compute per dollar spent than the US giants; the Seattle Times and Newsday sue OpenAI and Microsoft over copyright while the split of the Anthropic settlement pits authors against publishers; Jakub Pachocki, OpenAI's chief scientist, describes an alien mind and calls for guardrails; and US job seekers hide instructions in their resumes to trick recruiters' AI14 actualités · 8 min
OpenAI ships GPT-6 Astra as three assistants go dark14 actualités · 8 min
Washington backs OpenAI against the New York Times14 actualités · 9 min
OpenAI rates Astra at critical cyber capability14 actualités · 9 min
The European Commission ranks ChatGPT among its most closely watched services under the DSA, the first generative AI targeted, OpenAI moves some of its large customers to performance-based billing, Anthropic signs $35 billion in compute with Lambda ahead of its IPO, and Nvidia invests $3.5 billion in MediaTek to stay indispensable14 actualités · 8 min
Août 202610 éditions
Sony Music and Warner sue Anthropic for \"one of the largest thefts of intellectual property in history\" on the eve of its stock market listing, Bill Gates calls for taxing machines, Google rolls out its AI answers by default at the expense of the blue links, and Caterpillar applies thirty years of mining automation to AI14 actualités · 7 min
Nvidia moves on Hugging Face for $12.9 billion, the open-source platform founded by three French entrepreneurs, the same week it was breached by OpenAI's agents; a federal judge overturns the Pentagon's ban on Anthropic, OpenAI brings advertising to ChatGPT in India, Zhipu unveils an open model trained on 100,000 Chinese chips, and OpenAI documents what critical thinking changes about using a chatbot14 actualités · 7 min
Nvidia posts a record $96.2 billion in quarterly revenue and looks ahead to 2028, while Amazon triples its chip order and Anthropic signs $45 billion of compute with Nscale; on the same day, OpenAI releases its report on the Hugging Face breach carried out by nearly 700 of its own agents, Bill Gates breaks his silence to call for a robot tax, and the CNIL updates its open-model traceability tool14 actualités · 8 min
Open-weight models pass proprietary ones on Vercel14 actualités · 7 min
Uber fined €825 million over automated decisions13 actualités · 7 min
Nvidia warns its biggest customers of a price rise above 15% on its AI servers as memory costs soar; the same day SoftBank places a record bond to fund OpenAI, Samsung returns $79 billion under the pressure of the AI boom, Alibaba raises $10.2 billion in Hong Kong, and Hugging Face is said to be exploring a sale at around $13 billion14 actualités · 8 min
Anthropic eyes a SpaceX-sized listing as enterprise buyers switch models14 actualités · 7 min
The AI Act makes provenance a job14 actualités · 9 min
OpenAI freezes part of Astra over cyber risk14 actualités · 8 min
Jeff Dean leaves Google as DeepMind reshuffles its command13 actualités · 7 min

25 éditions publiées, 344 actualités traitées depuis le 6 August 2026. Voir toutes les publications de veille

Aller plus loin

Les sujets de cette veille, traités en profondeur

La régulation suivie dans cette veille est celle que nous appliquons dans nos projets. Voir les agents IA en entreprise. La rubrique se suit aussi par flux, sur Feedly, un connecteur Slack ou Teams, ou une automatisation maison : Flux RSS Notre méthode et nos sources sont détaillées dans la politique éditoriale.

Le fondateur
Mathias Nizan, fondateur de Masteria

Mathias Nizan

Fondateur de Masteria, cabinet de conseil et développement IA

Mathias Nizan a fondé Masteria en 2022 à Lyon. Cabinet spécialisé uniquement sur l'intelligence artificielle et indépendant des éditeurs, Masteria a formé plus de 1 500 professionnels et accompagne PME, ETI et grands groupes, du cadrage stratégique au développement des solutions sur mesure, en France, en Suisse et en Belgique.

« L'intelligence artificielle ne remplace pas les humains. Elle décuple leur potentiel. »

Spécialiste IA depuis 2022 +1 500 professionnels formés Cabinet indépendant des éditeurs Lyon · France · Suisse · Belgique LinkedIn

Former vos équipes sur ce que vous lisez ici

L'actualité suivie chaque matin nourrit nos formations et nos missions de conseil. Masteria accompagne dirigeants, chefs de projet, développeurs et juristes sur l'IA générative depuis 2022.

Parler de votre projet

Réponse sous 24 h · Organisme certifié Qualiopi · Lyon, France, Suisse, Belgique

On compte les visites, pas les visiteurs.

En acceptant, vous aidez un cabinet indépendant à savoir quelles pages vous sont utiles et à améliorer les autres. Aucune publicité, aucune revente, aucun suivi d'un site à l'autre, et vous pouvez changer d'avis à tout moment.