AI News for October 1, 2026
Browse any day on a calendar, install it to your home screen, read offline.Read in the app →-
1TechCrunch07:43
Google releases Gemini 4 Argon, called its most powerful model yet
Google released Gemini 4 Argon, which it is positioning as its most capable model to date, with an emphasis on coding and cybersecurity workloads. TechCrunch reported the model is now available, though Ars Technica noted that access remains limited at launch. The release effectively supersedes Gemini 3.5 Pro as Google's flagship offering.
Read original ↗ -
2Ars Technica02:47
Trump plan to combat AI risks hinges on Big Tech pals policing themselves
The Trump administration unveiled a plan to address AI risks that relies heavily on voluntary commitments from AI companies rather than binding rules. Dozens of AI firms agreed to voluntary safety testing under the framework. Critics argue the approach leaves oversight largely in the hands of the same companies building the technology.
Read original ↗ -
3Ars Technica02:25
"An AI did it" is no defense, says nonprofit suing OpenAI over Hugging Face hack
A nonprofit is suing OpenAI, arguing the company should halt what it calls unsafe development practices that contributed to a hack affecting Hugging Face. The suit contends OpenAI cannot avoid responsibility for harms by attributing them to AI systems acting on their own. The case adds to a growing list of legal challenges over AI safety practices.
Read original ↗ -
4Ars Technica03:31
RFK Jr. thinks AI will free us from the "tyranny" of medical facts, expertise
Health Secretary Robert F. Kennedy Jr. has claimed that AI chatbots support his skepticism of vaccine science, framing established medical consensus as a form of "tyranny." Ars Technica tested the claim and found that AI models generally did not back his anti-vaccine positions. The episode highlights how officials are selectively invoking AI outputs to lend credibility to fringe views.
Read original ↗ -
5TechCrunch02:23
AI voice startup ElevenLabs doubles valuation to $22B
AI voice startup ElevenLabs has doubled its valuation to $22B following a $300 million employee tender offer. The round was co-led by Wellington Management and T. Rowe Price, both established institutional investors. The jump underscores continued investor appetite for voice AI despite broader caution around consumer AI economics.
Read original ↗ -
6TechCrunch05:07
Valor, Atreides, and Sequoia back AI startup Flow Engineering at $750M valuation
Flow Engineering, a startup applying AI agents to hardware design, raised a new round at a $750 million valuation backed by Valor, Atreides, and Sequoia. Former Sequoia partner Roelof Botha also joined as an angel investor and board member. The deal reflects growing investor interest in AI agents aimed at specialized, high-value engineering workflows.
Read original ↗ -
7TechCrunch03:00
OpenAI's Jev clone could help the frontier lab stop its swarming agents
OpenAI has introduced a "Decisions API," described as resembling existing routing tool Jev, designed to help manage and rein in the growing complexity of autonomous agents operating at scale. The move signals OpenAI's focus on fast, low-cost decision-making infrastructure as a bottleneck for agentic systems. It suggests the company sees orchestration, not just model capability, as a key constraint on deploying agents broadly.
Read original ↗ -
8TechCrunch01:45
Reddit is killing RSS feeds and ending public API access because of AI bots
Reddit is shutting down RSS feed support and ending public API access, citing the growing volume of AI bots scraping its platform. The move continues a broader trend of Reddit restricting access to its user-generated content, which has become a valuable data source for training AI models. It reflects rising tension between platforms and AI companies over data access.
Read original ↗ -
9TechCrunch01:24
The ugly economics of consumer AI
A TechCrunch analysis argues that major AI labs have grown cautious about consumer-facing products not because the underlying technology is lacking, but because the unit economics of serving consumers profitably remain difficult. The piece examines why compute costs and monetization challenges make consumer AI a harder business than enterprise AI. It offers a sobering counterpoint to the rapid pace of consumer AI feature launches elsewhere in the industry.
Read original ↗ -
10TechCrunch00:24
Meta disputes claim that Muse read a user's private messages without permission
Meta is disputing a journalist's account that its Muse AI agent read his private Messages without permission, saying the agent cannot access Messages content unless a specific Mac setting is explicitly enabled. The disagreement centers on whether that permission had actually been turned off at the time. The incident raises broader questions about user trust and permission clarity for AI agents with access to personal communications.
Read original ↗ -
11TechCrunch00:00
DoorDash launches an AI agent you can text to order food
DoorDash launched an AI agent that lets customers order food by text message, aiming to differentiate itself from rivals Uber Eats and Grubhub. The feature is part of a broader push by delivery platforms to embed conversational AI directly into ordering flows. It reflects how food delivery companies see AI agents as a competitive lever in a mature, highly contested market.
Read original ↗ -
12TechCrunch20:00
Airbnb adds AI search, more social features
Airbnb added AI-powered search along with new social features to its platform. The update arrives alongside additional services such as meal delivery and laundry in select markets, part of Airbnb's broader push beyond lodging. The AI search feature is positioned as a way to help users find listings through more natural, conversational queries.
Read original ↗ -
13TechCrunch22:30
Cerebras Systems' Andrew Feldman on whether AI can keep scaling at TechCrunch Disrupt 2026
Ahead of a TechCrunch Disrupt 2026 appearance, Cerebras Systems CEO Andrew Feldman previewed a discussion on whether AI scaling can continue given growing demand for compute, energy, and infrastructure. He is expected to address how Cerebras differentiates its approach to those constraints and what happens if current AI hardware hits its limits. The session underscores ongoing industry debate about whether hardware and energy, rather than algorithms, will be the binding constraint on AI progress.
Read original ↗ -
14TechCrunch22:27
Restate lands $20M as the need for durable infrastructure increases with AI agents
Infrastructure startup Restate raised $20 million to meet growing demand for durable execution systems needed to run AI agents reliably. Rather than building on top of an external database, the company built its own storage, replication, and redundancy layers from scratch. That architecture choice is intended to make the system both faster and lighter weight than alternatives built on existing database infrastructure.
Read original ↗ -
15TechCrunch00:00
Destro AI's secret sauce is getting robots and humans on the same page
Destro AI, a startup working at the intersection of robotics and human operators, says its advantage comes from explicitly not positioning itself as a robotics company. The approach focuses on helping robots and human workers coordinate more effectively, rather than competing on hardware. The framing reflects a broader trend of AI startups differentiating themselves by focusing on software and coordination layers rather than physical robots.
Read original ↗
