★ Top Story
TechCrunch AIAnthropic has proposed AI safety legislation that would allow regulators to block a model's release if it poses a risk of killing 50 or more people — but the fine print is notable. The bill would actually require federal agencies to obtain a judge's permission before halting any model, limiting the government's ability to act unilaterally. Critics may see this as safety theater that protects AI labs more than the public.
Read the original →
TechCrunch AIThe U.S. government has directed all federal civilian agencies to purge DeepSeek products and any third-party services built on its technology, citing national security concerns. The order extends beyond direct DeepSeek apps to cover downstream integrations, significantly broadening its scope. The move marks an escalation in Washington's effort to limit Chinese AI influence within government infrastructure.
Read more →
TechCrunch AIAnthropic is reportedly struggling to lock down sufficient compute from major cloud providers, many of which now view the AI lab as a direct competitor rather than a customer. This tension is emerging at a particularly awkward moment, as Anthropic's infrastructure spending is ballooning ahead of a planned November IPO. The situation raises questions about whether Anthropic can scale its operations without more reliable and neutral compute partners.
Read more →
TechCrunch AIAnthropic is reportedly in advanced discussions with compute provider Nscale for a $45 billion infrastructure deal, which would be the latest in a string of massive compute commitments the company has made over the past year. The talks underscore just how capital-intensive frontier AI development has become, and how aggressively Anthropic is securing resources ahead of its IPO. If finalized, the deal would further cement Nscale as a major player in the AI infrastructure arms race.
Read more →
TechCrunch AISpaceX reported net income of $4.4 billion in Q3, with its AI division identified as the primary growth engine behind the company's profitability. The results exceeded analyst expectations and signal that SpaceX has successfully diversified beyond its core launch and satellite businesses. The strong AI performance adds a new dimension to SpaceX's valuation story as it continues to expand its commercial footprint.
Read more →
TechCrunch AIOpenAI has acquired biotech startup Orchestra Bio, folding its team and technology into Rosalind, OpenAI's AI model designed for life sciences research. The acquisition also bolsters the Rosalind Workbench, a tool aimed at helping researchers design experiments and interpret results. The move signals OpenAI's deepening ambitions in the biotech sector as it looks to apply its AI capabilities beyond consumer and enterprise software.
Read more →
The Verge AICoreWeave's bonds slipped into distressed territory last week, with investors demanding over 1,000 basis points more yield than government bonds — a threshold that signals serious financial stress. The development has drawn attention from prominent investors questioning whether the broader AI infrastructure buildout has become a private credit bubble. If CoreWeave's financial troubles deepen, the ripple effects could hit AI labs and cloud customers that depend on its GPU capacity.
Read more →
TechCrunch AIGoogle has released a developer tool that makes it easier for PyTorch users to migrate their AI applications away from Nvidia GPUs, joining Amazon in a coordinated push to erode Nvidia's dominance among AI developers. The tools lower the friction of switching to alternative hardware, which could gradually chip away at Nvidia's ecosystem lock-in. This represents a significant strategic alignment between two of the cloud industry's biggest players against a common competitor.
Read more →
The Verge AIMeta has temporarily disabled a Muse feature that allowed users to share AI conversations after security researchers at Claroc discovered more than 6,800 publicly accessible Muse chat links through public DNS records. The privacy lapse is particularly notable given that Muse reached 10 million U.S. downloads in just 40 days, meaning a large user base may have had conversations inadvertently exposed. Meta has not yet clarified how long the feature was leaking data or how many users were affected.
Read more →
TechCrunch AIMeta's AI assistant app Muse has reached 10 million U.S. downloads in just 40 days, a milestone achieved by only 11 other apps in history. Beyond raw download numbers, new data shows Muse is retaining users at higher rates than competing AI apps, suggesting strong product-market fit. The growth trajectory positions Muse as a serious contender in the consumer AI assistant market.
Read more →
Ars TechnicaMeta has quietly revised its commercial terms of service to remove explicit references to AI training, a move that follows significant user and creator backlash over data scraping practices. However, the company stopped short of adding language that would honor robots.txt directives, meaning it still reserves the right to scrape content from the web. Critics argue the change is largely cosmetic and does not meaningfully protect creators from having their work used to train AI models.
Read more →
Ars TechnicaA court has ordered Meta to identify the executives who approved the downloading of pirated books via torrents for use in AI training, after the company attempted to deflect blame onto lower-level employees. Meta's legal team argued that no single person formally "decided" to pirate the material, a framing the court rejected. The ruling could expose senior leadership to personal legal liability in ongoing copyright litigation.
Read more →
Ars TechnicaGoogle is expanding its Gemini Personal Intelligence feature to Android Auto and vehicles with Google built-in, allowing the in-car AI to access data from Gmail, Calendar, Maps, Photos, YouTube, and Search. The integration means drivers can ask Gemini context-aware questions — like checking a flight status while en route to the airport — without touching their phone. The rollout raises fresh questions about data privacy and prompts users to review their Gemini permissions.
Read more →
The Verge AISpotify has removed more than 50 million AI-generated or spam tracks from its recommendation algorithms over the past year, in a sweeping effort to protect human artists from being buried by low-quality synthetic content. The company also updated its recommendation system to surface more music made by human creators. The scale of the purge highlights how aggressively AI-generated content has flooded streaming platforms and the growing pressure on services to police it.
Read more →
TechCrunch AIMicrosoft is advising enterprises to monitor AI agents with the same vigilance applied to insider threats, arguing that frontier models cannot be fully trusted when operating within systems outside the direct control of AI labs. The warning reflects growing concern that autonomous agents — which can take actions across enterprise software — introduce novel security risks that traditional IT defenses weren't designed to handle. Microsoft's guidance signals a broader industry reckoning with agentic AI's security implications.
Read more →
TechCrunch AIEka Robotics has reached a $1.45 billion valuation on the strength of a robotic arm that experts say is genuinely capable of picking up a wide variety of objects with smooth, human-like motion — a longstanding challenge in industrial robotics. The company's physical AI model is drawing comparisons to a potential "ChatGPT moment" for robotics, where a single product demonstrates a step-change in capability. If the technology scales, it could accelerate automation across warehousing, manufacturing, and logistics.
Read more →
Google (blog.google)Google's annual e-Conomy SEA report projects Southeast Asia's digital economy will reach $368 billion in 2026, with AI adoption identified as a primary accelerant of the region's growth. The report forecasts the region could hit $1 trillion in digital economic value by 2035 if current trends hold. The findings underscore Southeast Asia's emergence as one of the world's most dynamic AI adoption markets.
Read more →
The Verge AIA new report finds the United States is constructing more fossil fuel power infrastructure than at any point in the past decade, driven largely by surging electricity demand from AI data centers, factories, and electric vehicles. Renewable energy growth is simultaneously slowing, creating a widening gap between clean energy ambitions and the reality of powering the AI buildout. The trend puts the tech industry's sustainability commitments under increasing scrutiny as its energy footprint expands rapidly.
Read more →No stories from this source today.