Actualités — 2026-09-15

Cathie Wood, founder of ARK Invest, called warnings of AI extinction 'ridiculous' at a Sydney briefing, praising the industry for openly discussing its risks. Wood's firm is heavily invested in AI companies, including Tesla, SpaceX, and OpenAI. Lire la suite

SoftBank Group shares dropped as much as 13% in Tokyo after AI-linked stocks sold off worldwide, sparked by calls from Anthropic and OpenAI to slow the development of more capable models. The decline came despite SoftBank securing an $11.87 billion loan to fund its OpenAI investment. Lire la suite

Bilal Chughtai, a former researcher at Google DeepMind, has expressed extreme concern about the potential dangers of artificial intelligence, warning that it could "kill us all" if left unchecked. Chughtai, who worked on AGI safety and alignment research until July, believes that coordination is needed to avoid a "manic race" between AI companies and to pace development at a speed that society can handle. Lire la suite

Salesforce has launched Koa, its first reasoning model, built on Nvidia's open-weight Nemotron model. Koa is designed to excel at sales, marketing, and customer-support-related tasks, offering an alternative to closed frontier models. It's trained on synthetic data, reducing AI spending and ensuring customer data security. Lire la suite

A new Mozilla report reveals that the performance gap between frontier AI models from US tech companies and open-weights models from Chinese companies has narrowed to just 4.4 months. This finding suggests that many companies are shifting to cheaper open models for routine work, while frontier models are worth the cost for specific workloads. Lire la suite

Kamala Harris is urging Congress to establish a new federal entity to oversee AI development and calling for a treaty with countries like China to address the dangers of AI. Her proposal comes as other high-profile figures, including Barack Obama and Pete Buttigieg, are also weighing in on the need for AI regulation. Lire la suite

A cross-party committee of MPs and peers is calling for a dedicated UK AI law and a single regulator with enforcement powers, with OpenAI backing binding rules for companies building the most powerful AI systems. Lire la suite

Anthropic is reportedly the unnamed customer behind a $13.7 billion, six-year computing contract with Rum Group, the parent company of video platform Rumble. The deal adds to Anthropic's growing list of compute agreements, including Google, SpaceX, and Nscale. Lire la suite

Over 5,000 mathematicians, including 25 Fields medalists, have signed a declaration expressing concerns about AI's impact on their field, citing the potential for AI to undercut human mathematicians' work and hinder conceptual understanding. Lire la suite

A team of researchers has introduced GRP-Obliteration, a method that can unalign safety-aligned AI models using a single unlabeled prompt, potentially undermining their robustness. The technique, which leverages Group Relative Policy Optimization, can reliably remove safety constraints from target models while preserving their utility. Lire la suite

More than 70 UK lawmakers are urging Prime Minister Andy Burnham to back a ban on building artificial superintelligence in the UK, citing the potential threat to national and international security. The proposed bill would prohibit the development, deployment, and operation of superintelligent AI systems, and give the government powers to monitor and control them. Lire la suite

Researchers at OpenShell are using formal methods to ensure AI agents operate within established permissions, reducing the risk of unintended actions. By applying formal verification techniques, they can prove that proposed policy changes do not exceed pre-approved limits. Lire la suite

OpenAI CEO Sam Altman outlined a plan for the AI industry to slow down and implement safety frameworks, stating that the company won't wait for Congress to act. Altman's proposal includes shared industry standards and independent auditors to ensure AI safety. Lire la suite

Temporal, a company that makes software to keep applications running when something fails, has raised $550 million in a Series E round that values it at $12.55 billion. The funding will go into research and development, sales, and global operations to support growing demand from businesses putting AI agents into production. Lire la suite

Anthropic has told shareholders it expects positive adjusted operating income for a second straight quarter, with gross margins above 80%. The company is on track for an IPO, aiming to list on Nasdaq as soon as October, and is finalizing a $15 billion revolving credit facility. Lire la suite

OpenAI's top strategist, Chris Lehane, calls the industry's turn to state-level AI lawmaking 'reverse federalism,' while the company's president helps fund a super PAC pushing for a national framework to override state rules. This comes as the industry navigates a complex landscape of regulation and governance. Lire la suite

Anthropic co-founder Jack Clark suggests lawmakers may need to make AI 'kill switches' mandatory to prevent potential risks. He also proposes outside parties should be able to verify these switches, sparking a wider policy debate on AI regulation. Lire la suite

Donald Trump has labeled fears about AI a 'hoax' in a series of posts on Truth Social, questioning why industry leaders want regulation that could drive them into bankruptcy. The US president's comments come as lawmakers and experts debate the need for oversight of the rapidly developing technology. Lire la suite

European Central Bank President Christine Lagarde warns that Europe's reliance on foreign AI could be used as leverage in trade negotiations, urging the continent to produce its own AI to maintain sovereignty and avoid being cut off from critical technology. Lire la suite

Meta introduces Meta One, a new subscription service offering expanded AI usage and premium features across Facebook, Instagram, and WhatsApp, following its $14.3 billion investment in Scale AI. Lire la suite

Researchers have introduced NüshuRescue, an AI-driven framework designed to preserve and revitalize the endangered Nüshu language, used by Yao women in China. The framework trains large language models on minimal data, automating evaluation and expanding target corpora to accelerate linguistic revitalization. Lire la suite

Grand View Research and MarketsandMarkets forecast the agentic AI market to reach $24.5 billion and $52.62 billion by 2030, respectively, with a compound growth rate of around 46%. The discrepancy in market size estimates highlights the varying definitions of agentic AI, while the convergence on growth rate signals a strong signal for the category's potential. Lire la suite

AI agents from startup iLands are sending waves of unsolicited messages to social media platforms and writers, offering to cite their work in exchange for a fee and sparking concerns over automation and the future of writing. Lire la suite

Top AI leaders are urging for regulations to control the rapid development of artificial intelligence, citing concerns over safety and the potential for AI to 'run away' without proper guardrails. The push for regulation comes as the US administration remains skeptical, with President Trump suggesting that a 'strong and smart' president is all the control AI needs. Lire la suite

A recent test pitted GPT-5.6 Luna against GPT-6 Astra in a code review challenge, with Luna finding 69 verified bugs across 50 pull requests, compared to Astra's 92. Despite the difference, Luna's cost was significantly lower, at $0.20 for the entire run, versus Astra's $5.66. Lire la suite

OpenAI has purchased Glass Imaging, a smartphone camera maker, in a deal worth over $300 million. The company, founded by former Apple engineers, uses AI to improve smartphone camera images. This acquisition comes as OpenAI is rumored to be working on its own hardware, including smartphones and AI companion devices. Lire la suite

Former FTC chair Lina Khan argues that existing laws can be used to hold AI companies and executives accountable for their actions, citing a 1934 Supreme Court precedent. Khan's comments come as AI leaders try to influence regulators and downplay the risks of their technology. Lire la suite

dbt Charts, a new open-source language, is poised to transform the way we interact with data. By moving charts out of BI tools and into code, dbt Charts enables seamless collaboration between humans and AI agents, making data analysis more efficient and accessible. Lire la suite