OpenAI shipped its first model rated Critical for cyber
OpenAI released GPT-6 Cyber on Thursday and classified it Critical under its Preparedness Framework, the top cyber tier, one notch above where GPT-5.6 Cyber sat. On OpenAI’s own ExploitGym evaluation it scored 48.5% against 32.1% for its predecessor. Access is restricted to vetted Daybreak Red customers through the API and Codex, with a wider Red rollout promised over the coming weeks.
OpenAI says the model surfaced 81 previously unknown high-severity vulnerabilities in Firefox, and The Register reports Mozilla has started shipping fixes. The companion writeup, A thousand fewer paths to root, describes an October campaign that analyzed 2,308 candidate Linux kernel findings, dynamically reproduced 1,724 of them inside 30 hours, and ended with 948 working local privilege escalation exploits, 588 of them previously unknown. Total bill: $1,213,430 in API credits, about $1,279 per exploit.
On scope discipline, OpenAI says the model stayed inside its authorization boundary over 99% of the time and that 0.07% of cases went out of scope. It declined to say what happened in those. Worth remembering that GPT-6.1 Astra was shelved in September for overstepping authorized boundaries, which is why the number is in the release notes at all.
Agility’s Robots Meet Wall Street’s Calculator
Agility Robotics, the proud parent of the Digit factory robot, skipped onto the Nasdaq yesterday under the ticker AGLT. That makes it the first American pure-play humanoid builder with actual commercial gigs to brave the public markets. A merger with blank-check company Churchill Capital XI and related cash drops stuffed roughly $465 million into Agility’s pockets before transaction fees.
Almost a year to the day after Figure AI’s CEO publicly bet Agility would go bankrupt, the Oregon startup has a shiny new listing. It also has a sobering reality check for investors: a paltry $1.78 million in 2025 net sales against a brutal $140.2 million operating loss. Now Wall Street gets to see what happens when sci-fi humanoid pitches collide with cold, hard income statements.
To its credit, Digit has already hauled over 100,000 totes for logistics giant GXO, which plans two additional US deployments in early 2027. Agility’s next-gen Digit 5 model is ditching the safety cages to mingle with human coworkers and targeting a caffeine-fueled 20-plus-hour workday via rapid charging. Early access is expected in the first half of 2027, with general availability targeted for year-end.
But don’t let the headline-grabbing $300 million in pre-orders fool you—that isn’t booked revenue. SEC filings reveal the figure is heavily tied to a 1,000-robot agreement with a single related-party customer and entirely conditional on the bots actually hitting promised milestones. That leaves manufacturers juggling two gambles: whether Digit can reliably do the heavy lifting, and whether Agility can financially survive long enough to deliver them.
Digit can carry the totes. Agility still has to carry the business.
Data centers lost 40 of 41 ballot measures
Tuesday’s midterms put data centers directly on the ballot 41 times across ten states, and voters backed a restriction or an outright ban in all but one. Ohio carried 21 of them: Conneaut, Defiance, Oregon and Sunbury capped data centers at 25 megawatts, while Granville, Hubbard, Urbana, Upper Sandusky and Piqua banned them outright. All seven Illinois measures went against development. A Wisconsin measure now requires a public vote on any development over $450 million. A Bowling Green State University poll last month had 71% of Ohioans backing a temporary construction ban, so this was not a surprise to anyone who was looking.
Markets caught up on Thursday. The Nasdaq fell 5.52% to 24,315, its worst day since April 2025 and into correction territory, more than 10% off its October 6 record. The S&P 500 dropped 4.36% to 7,353.73 and the Philadelphia Semiconductor Index lost more than 12%. Nvidia closed down 9.26% at $181.60, AMD fell 17.26%, Broadcom 10.43%, and the neocloud names IREN, Applied Digital and Nebius each shed more than 14%. eToro’s Jakub Rochlitz read it simply: the market now expects the buildout to be “slower and more expensive.”
OpenAI and Meta each bought a coding startup on the same day
OpenAI agreed on Wednesday to acquire Pierre Computer Company, which builds a Code Storage service plus the open-source Trees and Diffs libraries that Codex already uses for file tree navigation and visual diffing. Terms were not disclosed and the deal is subject to regulatory approval. The plan is to pull Code Storage into Codex, turning it from a completion tool into a repository and code review environment. Axios framed it as a run at GitHub, which tracks with Pando, the Git-compatible backend we covered Wednesday. OpenAI says the open-source projects stay available.
Hours later Meta acquired Conductor, founded in 2024, which runs multiple coding agents in parallel workspaces across cloud, local machines and phones using models from Anthropic, OpenAI and others. Terms undisclosed. Founder Charlie Holtz said he is looking forward to working with Alexandr Wang and Nat Friedman at Meta Superintelligence Labs, and teased something new in early 2027. If you use it: the cloud service shuts down November 20 and user cloud data gets deleted after that, with core features running until early next year.
Cat Wu, who led Claude Code product and left Anthropic on October 30, has joined OpenAI to run product management for Core Products, under a week after walking out the door.
Midterms trigger AI crackdown
Tuesday’s midterm elections handed Democrats control of Congress, signaling a sharp shift away from the Trump administration’s hands-off approach to AI regulation. Voters across the country backed restrictions on data center development, while incoming lawmakers are promising congressional hearings and even an AI development pause.
- Newly elected Senator Abdul El Sayed (D-MI) has called for a “full-on pause” on AI development, while Representative Yassamin Ansari (D-AZ) is advocating against AI superintelligence research.
- The election results triggered market volatility: tech stocks surged Wednesday before AI chipmakers Nvidia, Micron, and AMD led a sharp Thursday selloff as investors reassessed the regulatory outlook.
- Texas Governor Greg Abbott has proposed pausing some existing data center approvals, with analysts warning approvals could face six months of delays due to new regulatory red tape.
For enterprise AI buyers, the midterms create immediate uncertainty around infrastructure capacity and vendor stability. Companies relying on frontier model providers like OpenAI, Anthropic, and Google should expect potential policy changes and model release slowdowns as congressional investigations unfold. The regulatory patchwork between states is widening: an AI application acceptable in one state could require additional disclosure or auditing in another. Enterprises should treat AI governance as a core business practice and conduct regular risk assessments now.
Moonshot AI leased Nvidia Blackwell chips via European data centers to bypass U.S. export controls
Chinese AI lab Moonshot AI pursued access to Nvidia’s restricted Blackwell GPUs through data centers in Europe after the White House flagged it for accessing advanced chips remotely in Thailand. Sources told CNBC that Moonshot sought to lease capacity on clusters of 8,000+ Blackwell GPUs for inference and training, with talks continuing even after U.S. officials publicly raised concerns in July.
U.S. export controls ban direct shipments of Nvidia’s most advanced chips to China, but current law doesn’t restrict remote access via overseas cloud services, creating a compliance loophole that Chinese AI labs have aggressively exploited. Moonshot reportedly offered to pay 30-50% more upfront than non-Chinese customers to secure deals with European cloud providers.
Chinese AI labs are gaining ground on U.S. competitors while Washington debates closing the remote access loophole. The Remote Access Security Act passed the House in January but hasn’t advanced in the Senate, leaving regulatory gaps that allow Chinese developers to access cutting-edge compute the U.S. intended to restrict.
Nvidia Eyes Old Telecom Sites
Nvidia has approached Apollo about funding a network of small AI data centers across 20 states, retrofitting old telecom buildings that already have power, fiber and zoning. The plan would push AI compute to the edge and unlock stranded real estate as demand for inference capacity spreads.
Zipline Raises $1.5B at $21.5B
Drone delivery operator Zipline raised $1.5 billion at a $21.5 billion valuation as Walmart expands its bet on the service. The round makes Zipline one of the most valuable robotics companies in the country and signals retail giants now treat autonomous delivery as core infrastructure rather than an experiment.
China Blocks ABB Deal
Chinese regulators have blocked SoftBank’s $5.4 billion acquisition of ABB’s robotics division, citing concerns that robot data could move overseas and train foreign AI models. The veto injects a new data-sovereignty risk into global robotics M&A and complicates exit paths for Western industrial automation assets.
Humble Robotics Raises $51 Million
Autonomous trucking startup Humble Robotics raised $51 million to pilot cabless haulers. The company is removing the driver’s cab entirely and running freight pilots on closed routes, betting that purpose-built driverless trucks can undercut human-driven fleets on cost long before full highway autonomy arrives.
Chip Slump Drags Nasdaq Lower
Nvidia and Qualcomm both fell more than 9% as investors dumped AI-buildout stocks on fears the infrastructure spending cycle slows after the midterm elections. The Nasdaq logged its worst day since April, a stress test for the valuations underpinning the entire AI capital expenditure thesis.
AppLovin Posts $2.18B
AppLovin reported $2.18 billion in third-quarter revenue, up about 55% year over year, and $1.44 billion in net income, a net margin of roughly 66%. The results topped expectations, but investors remain cautious about how durable that margin is as ad-tech competition intensifies.
Intel Joins Hark’s $700M Round
Intel joined a $700 million funding round for AI startup Hark alongside Nvidia and AMD Ventures, a rare moment of three rival chipmakers backing the same company. Hark builds AI-powered industrial inspection, and the strategic overlap suggests chip vendors now court application-layer startups to drive silicon demand.
Varda Lands $251M Series D
Space manufacturing company Varda raised a $251 million Series D to scale in-space drug production. The company flies capsules that crystallize pharmaceuticals in microgravity, and the new capital funds more missions as it works to turn orbital factories from demos into a repeatable commercial business.
New Glenn Slips to 2027
Blue Origin’s next New Glenn launch slipped to January 2027 as repairs from a May launch failure drag on. The delay hands SpaceX more breathing room in the heavy-lift race and tests Amazon’s Kuiper constellation timeline.
IBM, AMD Tackle Quantum Errors
IBM, AMD and Oak Ridge National Laboratory launched a quantum error-correction platform that pairs quantum processors with classical HPC. Error correction is the gate to useful quantum computing, and hybrid quantum-classical architectures are emerging as the leading approach to getting there faster.
OpenAI Lands in Apple Studio
Apple released its Creator Studio updates, putting OpenAI models beside Apple Intelligence inside Final Cut Pro, Logic Pro and the iWork apps. Final Cut gains built-in MotionVFX effects, and Keynote, Pages and Numbers add MCP support so external AI agents can edit documents directly.
Agility Robotics’ Muted Nasdaq Debut
Humanoid robot maker Agility Robotics closed its first Nasdaq session at $11.49, up just 2% from its offering price. After the marketing around the first humanoid IPO, the muted debut suggests public investors want revenue proof before paying up for embodied AI exposure.
Tesla Wants Cybercab Answers Hidden
Tesla responded to NHTSA’s special order demanding sworn proof that the Cybercab meets federal safety standards, then asked the agency to keep the entire response confidential. The secrecy request lands as the robotaxi ramps toward production and keeps regulators’ concerns out of public view.
Warby Leaks Google XR Glasses
Warby Parker appears to have prematurely revealed Google’s Android XR smart glasses, with a listing suggesting the Intelligent Eyewear launch is about a week away. The partnership puts Google’s XR platform directly against Meta’s Ray-Ban line this holiday season.
Perplexity Launches Teammates AI Agents
Perplexity launched Teammates, always-on AI agents for work that can hold multiple roles inside a workspace and pull in existing projects. The move puts the search startup in direct competition with OpenAI and Anthropic for the emerging agentic productivity market.