During a federal court testimony, Elon Musk seemingly admitted that his AI lab, xAI, utilized OpenAI's models for training, intensifying his legal battle against the ChatGPT maker. This revelation highlights the contentious practice of "distillation" and the escalating competition in the AI sector.
Pop superstar Taylor Swift has filed trademark applications to protect her image and voice, a move that underscores the escalating threat of AI deepfakes used in fraudulent online advertisements. A recent report revealed Swift and other celebrities are being impersonated in scam TikTok ads promoting fake services.
AI assistant maker Anthropic is reportedly on the verge of securing a massive $50 billion funding round, potentially valuing the company at an astounding $900 billion. This comes amid intense investor interest and rapid revenue growth, positioning it as a dominant force in the AI landscape.
Microsoft announced that its AI-powered Copilot now has over 20 million paid enterprise users, with CEO Satya Nadella citing significant engagement levels comparable to Outlook. This growth, driven by major enterprise adoptions and features like 'Agent mode,' defies earlier skepticism about the tool's usage.
A new decision-aware machine learning framework significantly increased access to essential medicines in Sierra Leone, boosting consumption by 19% in treated districts. The system, scaled nationwide, now covers an estimated two million women and children under five, demonstrating AI's potential to improve global health efficiency at low cost.
A landmark trial between Elon Musk and OpenAI CEO Sam Altman is underway, with Musk alleging deception over OpenAI's pivot from a nonprofit to a for-profit model. The outcome could determine OpenAI's future, leadership, and its impending IPO, potentially reshaping the entire AI industry.
Google has reportedly signed a classified deal allowing the US Pentagon to use its AI models for "any lawful government purpose," despite recent employee protests. This agreement places Google alongside other major AI firms, but raises questions about the company's ability to control the military's use of its technology.
A years-long legal feud between Elon Musk and OpenAI CEO Sam Altman culminates this week in a Northern California courtroom, potentially reshaping the future of the AI giant. The trial will delve into OpenAI's controversial pivot from a non-profit mission to a for-profit enterprise.
Microsoft and OpenAI have officially terminated their famed AGI agreement, making their partnership less exclusive and allowing OpenAI to use other cloud providers. This strategic shift reflects OpenAI's push for profitability and expansion into the enterprise market.
David Silver, a former DeepMind researcher, has raised $1.1 billion for his new AI lab, Ineffable Intelligence, to build a "superlearner" that learns without human data. This venture aims to revolutionize AI by leveraging reinforcement learning, a technique Silver mastered at DeepMind.
Chinese AI firm DeepSeek unveils its V4 model, matching leading rivals and optimizing for Huawei chips, while researchers push for "world models" to bridge the gap between digital and physical AI. This marks a new phase in the intense global AI competition.
Canadian AI startup Cohere is acquiring Germany-based Aleph Alpha to create a "sovereign AI" alternative, challenging US tech giants. The merger, backed by governments and Germany's Schwarz Group, aims to offer full data control for enterprises in regulated sectors.
Anthropic's "Project Deal" successfully demonstrated AI agents conducting real-world commerce, revealing that advanced models lead to better outcomes but also raising concerns about users' awareness of 'agent quality gaps'.
Thinking Machines Lab is aggressively recruiting top AI talent from Meta and other tech giants, fueled by a multi-billion dollar Google Cloud deal for Nvidia's latest chips and a $12 billion valuation. This signals a fierce competition for AI expertise in the rapidly evolving tech landscape.
Isomorphic Labs, a Google DeepMind spin-off, is preparing to launch human trials for drugs designed using its Nobel Prize-winning AI technology, marking a significant leap in AI-powered drug discovery. This move from computational prediction to clinical evaluation highlights AI's transformative potential in medicine.
Two Yale University students have secured $5.1 million in pre-seed funding to launch Series, an AI-powered social network that operates entirely within Apple's iMessage, aiming to facilitate 'warm connections' through a unique conversational interface.
ComfyUI, a startup providing granular control over AI-generated media, has achieved a $500 million valuation after securing a $30 million funding round. The platform addresses the limitations of prompt-based AI models, empowering creators with precise control over their outputs.
In a world saturated with AI news and hype, 'The Download' offers a crucial guide: '10 Things That Matter in AI Right Now.' This article cuts through the noise to identify the truly significant technological advancements, ethical debates, and real-world implications shaping the future of artificial intelligence.
The cyber capabilities of AI models have experts rattled. AI’s social skills may be just as dangerous.
Introducing Ace, the groundbreaking table tennis robot capable of defeating elite human players, leveraging a sophisticated high-speed perception system to predict complex ball movements with unparalleled accuracy, marking a revolution in sports and AI.
OpenAI and Infosys have formed a strategic partnership to integrate AI tools into Infosys's services, aiming to modernize software development, automate workflows, and deploy AI systems for businesses, with an initial focus on software engineering, legacy modernization, and DevOps.
North Korean hackers, once considered mediocre, are now leveraging AI tools for everything from "vibe coding" malware to creating fake websites, enabling them to steal millions. One group alone pilfered $12 million in three months, highlighting AI's role in lowering the skill barrier for state-sponsored cybercrime.
A recent Nature study reveals a critical flaw in current large language model (LLM) evaluation methods: focusing purely on accuracy can inadvertently incentivize models to hallucinate, generating convincing but false information to satisfy metrics.
