AI Pulse: Daily Digest — April 30, 2026

TechCrunch AI

Anthropic, the maker of the Claude AI assistant, has reportedly received multiple unsolicited investment offers valuing the company between $850 billion and $900 billion. If completed, a $50 billion raise at that valuation would make Anthropic one of the most valuable private companies in history, underscoring the extraordinary capital flowing into frontier AI labs.

The Verge

Elon Musk completed a second day of testimony in his lawsuit against OpenAI, where cross-examination repeatedly surfaced his own past statements and tweets as contradictory evidence. Observers noted that Musk's performance on the stand appeared to generate unexpected sympathy for OpenAI CEO Sam Altman, whose legal team has been methodically using Musk's public record against him.

The Verge

The Musk v. Altman trial has begun surfacing a trove of early OpenAI documents, including internal emails and corporate records from before the organization even had a name. The exhibits paint a detailed picture of the founding relationships and agreements at the heart of Musk's claim that OpenAI abandoned its nonprofit mission in favor of profit.

TechCrunch AI

Google Cloud crossed $20 billion in quarterly revenue for the first time, driven by surging enterprise demand for AI workloads. Despite the milestone, Alphabet executives acknowledged that infrastructure capacity constraints prevented even faster growth, signaling that data center buildout remains a critical bottleneck across the industry.

TechCrunch AI

Amazon Web Services delivered stronger-than-expected quarterly results, with AI-driven cloud demand fueling significant revenue growth. CEO Andy Jassy cautioned that capital expenditures will remain elevated in the near term as Amazon races to expand data center capacity to meet demand.

TechCrunch AI

Microsoft disclosed it now has more than 20 million paid Copilot users, pushing back against the narrative that enterprise AI assistants go largely unused. CEO Satya Nadella also signaled Microsoft intends to fully capitalize on its restructured deal with OpenAI, which gives Microsoft the right to resell OpenAI's technology to cloud customers without paying for it.

The Verge

New data from Sensor Tower shows ChatGPT app downloads are decelerating sharply, with uninstall rates surging 132 percent year-over-year in April and even higher in March. The slowdown raises questions about user retention and growth trajectory at a sensitive moment as OpenAI prepares for a potential public offering.

TechCrunch AI

Just one day after OpenAI renegotiated its exclusivity arrangement with Microsoft, Amazon announced it would offer a range of OpenAI models on AWS, including a new agent service. The rapid move illustrates how OpenAI is aggressively expanding its distribution across multiple cloud platforms simultaneously.

TechCrunch AI

SoftBank is spinning up a new robotics company with a novel mandate: use AI-powered robots to construct and operate the very data centers that AI requires. The venture is already reportedly eyeing a $100 billion IPO, reflecting SoftBank's bet that physical AI infrastructure will be as valuable as the software running on top of it.

TechCrunch AI

Google has signed a new contract to expand the U.S. Department of Defense's access to its AI tools, stepping in after Anthropic declined to allow its models to be used for domestic mass surveillance and autonomous weapons systems. The deal highlights the growing divergence among AI labs over where to draw ethical lines on military applications.

TechCrunch AI

Scout AI has secured $100 million to develop AI agents designed to help individual soldiers command fleets of autonomous vehicles in combat scenarios. TechCrunch visited the company's training facility, where real-world military environments are used to stress-test the AI systems before deployment.

Ars Technica

Seven families affected by the Tumbler Ridge school shooting in Canada have filed lawsuits against OpenAI and CEO Sam Altman, alleging the company failed to alert authorities after its systems flagged the suspected shooter's violent ChatGPT activity. Plaintiffs' attorneys claim OpenAI stayed silent to protect Altman's reputation and the company's IPO prospects.

Ars Technica

GitHub announced it will move Copilot to a consumption-based billing model, citing unsustainable inference costs generated by its heaviest users under the current flat-rate subscription. The change signals a broader industry reckoning with the economics of AI-powered developer tools as usage scales.

Hugging Face

A new analysis from Hugging Face argues that running rigorous evaluations on large AI models is rapidly becoming as expensive and time-consuming as training them, creating a hidden bottleneck in the development pipeline. As models grow more capable and evaluation suites more comprehensive, the cost of knowing whether a model is actually good is itself becoming a major constraint.

NVIDIA Blog

NVIDIA unveiled Nemotron 3 Nano Omni, an open multimodal model that combines vision, speech, and language processing into a single system rather than chaining separate specialized models together. NVIDIA claims the unified architecture enables AI agents to respond up to nine times more efficiently by eliminating the latency and context loss of passing data between discrete models.

The Verge

Authentication firm Copyleaks has documented a wave of AI-generated deepfake videos impersonating celebrities including Taylor Swift and Rihanna to promote fraudulent reward programs on TikTok. The scam ads typically splice AI-manipulated audio and video into real interview footage, making them difficult for casual viewers to identify as fake.

The Verge

Canonical's announcement that it will integrate AI features into Ubuntu has triggered a vocal backlash from Linux users who want the ability to opt out entirely, with some threatening to switch distributions or stay on older releases. The controversy reflects a broader tension in the open-source community between AI feature expansion and user control over their own systems.

Ars Technica

Japan Airlines has launched a pilot program deploying humanoid robots to handle baggage sorting at Haneda Airport, with an eye toward eventually using them for cargo loading and aircraft cabin cleaning. The trial is driven by Japan's acute labor shortage and represents one of the most visible real-world deployments of humanoid robotics in commercial aviation.