|
A hacker broke into John Deere and SharkNinja with an army of AI agents
A hacker used AI agents operating largely autonomously to breach more than 100 websites, including major companies like John Deere, SharkNinja, and Cummins, according to AI security firm Norforge. The incident highlights growing concerns about AI-powered cyberattacks and the ability of agentic systems to carry out sophisticated hacking campaigns with minimal human oversight.
|
|
Microsoft says AI agents are like insider threats as 'frontier models can't be trusted'
Microsoft is warning enterprises to treat AI agents as potential insider threats, particularly when those agents interact with systems not controlled by major AI labs like OpenAI and Anthropic. The company's stance reflects growing concern that frontier AI models are not yet reliable or secure enough to operate autonomously in sensitive corporate environments.
|
|
Hackers hit AI coding tool OpenCode, running up charges on users' AI accounts
Hackers breached part of OpenCode's online platform and stole users' API keys, resulting in unauthorized charges on victims' AI accounts. The open-source AI coding agent, seen as a rival to Claude Code and Cursor, found that the AI tools it relied on refused to analyze the malicious code during the incident.
|
|
Nscale and Anthropic are reportedly in talks for yet another $45B compute deal
Anthropic is reportedly in discussions with Nscale for a $45 billion compute deal, which would be the latest in a series of massive infrastructure agreements the AI company has struck over the past year. The deal underscores the enormous capital demands of frontier AI development and the race to secure dedicated compute capacity.
|
|
The fossil fuel comeback is here
A new report finds the US is building more fossil fuel infrastructure than it has in a decade, even as electricity demand climbs at its fastest pace in over ten years, driven by data centers, factories, and EVs. Despite surging demand, renewable energy growth is slowing, raising questions about the sustainability and environmental impact of the AI-fueled power boom.
|
|
CoreWeave's debt is becoming the whole AI industry's problem
CoreWeave's bonds slipped into distressed territory last week, with investors demanding 1,000 basis points more in yield than government bonds — a common definition of financial distress. The situation has prompted prominent investors like Chamath Palihapitiya to question whether the broader AI infrastructure buildout has become a private credit bubble.
|
|
Tolan snags $30M for its AI companion that helps users navigate life
AI companion app Tolan has raised an additional $30 million in funding after surpassing 250,000 active users. The app positions itself as a life-navigation tool, reflecting continued investor appetite for consumer-facing AI companion products.
|
|
Replika's founder built the AI assistant she pitched at Disrupt 12 years ago
Eugenia Kuyda, founder of AI companion app Replika, has launched a new startup called Wabi — a free, downloadable personal AI assistant that realizes a vision she first pitched at TechCrunch Disrupt more than a decade ago. The launch marks a full-circle moment for one of the AI companion space's most prominent founders.
|
|
Tenstorrent hits $6.2B valuation to scale open-weight AI chips
Nvidia challenger Tenstorrent raised $1.5 billion in new funding, pushing its valuation to $6.2 billion and allowing it to remain independent after facing acquisition interest earlier this year. The company is focused on scaling open-weight AI chips as an alternative to Nvidia's dominant hardware ecosystem.
|
|
Amazon is taking its biggest shot at Nvidia yet with new TorchNeuron dev tool
Amazon has launched TorchNeuron, a new developer tool aimed at breaking Nvidia's grip on PyTorch — the widely used open-source toolkit underlying most AI applications. The move is designed to make it easier for developers to run AI workloads on alternative chips, including Amazon's own silicon.
|
|
Nvidia launches new open models to run agents in the enterprise and on PCs
Nvidia has released a new set of open models targeting enterprise companies that want to run AI agents within their own data centers, as well as on-device private AI agents for PCs. The launch reflects Nvidia's push beyond hardware into the software and model layer of the AI stack.
|
|
Democrats think cracking down on AI could give them an edge
With polls predicting Democrats could flip at least one chamber of Congress in the midterm elections, the party is eyeing AI regulation as a potential political winner heading into November. A Democratic majority would put committee leadership back in their hands and give them the ability to shape the country's approach to one of its most contentious emerging technology issues.
|
|
Mamdani says feds have abdicated on AI, urges New Yorkers to push back
New York City Mayor Zohran Mamdani called on New Yorkers to resist the notion that an AI takeover is inevitable, arguing that federal authorities have failed to step up with meaningful oversight. He urged residents to demand regulations that protect the public and return power to communities rather than to technology companies.
|
|
Embedded AI evaluators won't be an easy solution, Hugging Face's policy chief says
The concept of embedding neutral third-party evaluators inside AI organizations has gained traction as a potential safety mechanism, but Hugging Face's policy chief is skeptical it will be easy to implement effectively. The debate comes as AI labs face mounting pressure following a series of safety incidents that have raised public concern about the risks of advanced AI systems.
|
|
JB Pritzker thinks AI could help us live to 100 — or threaten us like nukes
Illinois Governor JB Pritzker shared wide-ranging views on AI in a podcast interview, arguing the technology holds transformative potential for human longevity while also posing existential risks comparable to nuclear weapons. He also touched on the intersection of AI with privacy and data center policy, themes he sees driving a significant political realignment ahead of the midterms.
|
|
Laplace Intelligence red-teams robots and self-driving cars before they hit the real world
Laplace Intelligence, a University of Michigan spinout, uses AI agents and interactive simulation to stress-test autonomous robots and self-driving vehicles before they are deployed in real environments. The profitable startup is competing in Startup Battlefield at TechCrunch Disrupt and positions itself as a critical safety layer for physical AI systems.
|
|
People are using AI to make their smart TVs dumb again
Frustrated by ads, tracking, and sluggish interfaces on their smart TVs, some users are turning to AI agents to strip away bloatware and restore simpler, more controlled TV experiences. The trend highlights a growing consumer backlash against over-engineered smart home devices and a creative new use case for agentic AI.
|