Research archive
Browse source-linked technology intelligence across all five sections. Search by topic, company, or issue.
India's NVS-03 navigation satellite completes its first orbit-raising maneuver
ISRO reports that NVS-03 reached its intended transfer orbit and completed an initial engine burn after the October 30 launch. Navigation service still depends on further maneuvers and checkout.
Foxconn moves humanoid robots from factory plans into limited Texas deployment
Foxconn says more than ten humanoid robots are in internal use at Texas facilities, while its October technology showcase highlighted wheeled industrial designs. Production-scale gains remain unverified.
Tampered Claude Code extension exposes the risks of trusted developer marketplaces
Cursor disclosed a malicious Claude Code-branded extension distributed through its marketplace. The injected code attempted SSH persistence, and replacing the extension alone may not remove it.
NASA asks industry how to reboost, service or safely retire the Hubble telescope
NASA has issued an information request exploring commercial Hubble servicing, orbital reboosting and disposal. This is an assessment of options, not a funded rescue mission.
Google's orbital TPU returns its first AI response, with reliability tests still ahead
A Project Suncatcher prototype ran a Gemma 4 prompt on an orbiting Tensor Processing Unit and returned text to Earth. It is an initial functional demonstration, not an orbital data center.
Cornell maps coexisting superconducting states without proving a topological material
New uranium ditelluride measurements reveal a region where two superconducting states coexist. The finding refines the search for topological superconductivity but does not establish a working quantum device.
Walmart expands Symbotic warehouse automation under a roughly $6 billion agreement
A newly expanded multiyear agreement will extend Symbotic's automation capabilities to 25 U.S. Walmart distribution centers, with measurable operating gains still to be demonstrated.
OpenAI reorganizes its Model Spec around human intent, oversight and five values
OpenAI's October 29 Model Spec revision makes human intent and agency more explicit, while acknowledging that deployed models do not yet fully reflect the updated guidance.
ISS crew installs seventh roll-out solar array during milestone 100th U.S. spacewalk
NASA and Canadian Space Agency astronauts completed a seven-hour station power upgrade on October 29, with one final roll-out array still planned for November.
Alpine's GitLab incident reveals why a server upgrade may leave Ruby flaws exposed
Alpine Linux disclosed attempted exploitation of a critical GitLab flaw and found that upgrading its self-built server had not updated the vulnerable Ruby component.
Perseverance finds ultrafine Mars dust that could shape future spacesuits
Measurements of Martian dust as small as roughly 100 nanometers reveal how particles adhere to different spacesuit materials, informing future crew-protection designs.
Molecular qubits offer a possible chip-making route, but scaling remains unproven
NVision's peer-reviewed single-molecule spin-photon experiment and a new chip partnership suggest a manufacturing approach to quantum hardware, not a finished processor.
Gemini 4 Carbon raises cybersecurity benchmarks while limiting early access
Google's new Gemini 4 Carbon reports stronger vulnerability discovery and patching results, but initial access is restricted to selected cybersecurity defenders.
Swift's final observations close a two-decade mission after an unsuccessful orbital rescue
NASA ended the Swift observatory's science mission after orbital decay and a failed servicing attempt, leaving a large archive of gamma-ray burst observations.
Genie 4 makes AI-generated worlds interactive, but real-world fidelity remains unproven
Google DeepMind's Genie 4 adds live scene editing and generated audio to interactive world simulations, while short duration and visual drift limit current uses.
Unconventional AI adds NeuReality expertise to its efficiency-focused chip effort
A proposed acquisition brings experienced AI inference engineers to an experimental computing startup; its thousandfold efficiency target remains unproven.
NASA aims to shorten science-mission selection delays with a simpler competition model
NASA's Competition Model 2.0 targets faster decisions and lighter initial proposals, although reduced selection times do not automatically mean faster launches.
Cadence brings specialized AI models into the chip-design workflow
Cadence's new Maestro portfolio combines domain-trained AI models with chip-design software, but early efficiency claims still require independent validation.
Mitsubishi tests a humanoid robot workflow around CNC machine tools
Mitsubishi Electric's new machining system coordinates a humanoid robot and CNC equipment through an AI task planner, with a trade-show demonstration planned rather than proven factory-wide deployment.
Lumina simulation links the early universe's first stars to two waves of reionization
A newly published cosmological simulation models the first two billion years of cosmic evolution, following how early stars and growing black holes changed intergalactic hydrogen and helium.
CISA’s 2030 plan shifts the cyber-defense focus toward keeping critical services running
A new U.S. cyber strategy lays out three priorities—defend, harden and prevail—while promising a review of agency capabilities and a greater emphasis on measurable critical-infrastructure resilience.
Anthropic’s public IPO filing puts AI growth and infrastructure costs under scrutiny
Anthropic has publicly filed for a proposed Nasdaq listing, offering a new view of rapid AI-company growth alongside substantial infrastructure expenses, accounting losses and unresolved technology risks.
Texas disables automated permitting as 17 data-center approvals are suspended
Texas regulators suspended at least 17 data-center-related stormwater authorizations and shut an automated filing system while staff review applications against a statewide permitting pause.
Rocket Lab's latest HawkEye 360 launch expands space-based radio-frequency sensing
Three HawkEye 360 satellites reached low Earth orbit aboard Electron, extending a commercial system designed to locate and characterize radio emissions for security and maritime applications.
Oracle's emergency Java alert puts certificate trust back on the patching agenda
Oracle issued an out-of-cycle Java and GraalVM security alert tied to eMudhra root certificates, highlighting why trust-store updates need the same attention as application patches.
OpenAI and Poland's NASK agree to explore joint AI safety research
A new agreement pairs OpenAI with Poland's NASK research institute to explore AI evaluation and safeguards, but measurable outcomes from the new collaboration remain ahead.
Artemis III's next challenge is docking Orion with two different lunar landers
NASA's October mission outline describes an Earth-orbit test of Blue Origin and SpaceX lunar lander systems before a planned crewed lunar landing in 2028.
Google's AI personalization now has a clearer advertising privacy trade-off
Google's updated privacy policy permits some information voluntarily shared from Gmail, Drive and Photos with other services to inform personalized advertising, depending on account settings.
Actively exploited Arista VeloCloud flaw threatens on-premises network management
Arista confirms exploitation of a critical SQL injection vulnerability in on-premises VeloCloud Orchestrator, and CISA has added the flaw to its Known Exploited Vulnerabilities catalog.
Plex server owners: four exploited flaws move from theoretical to urgent
A vulnerability is different from an actively exploited vulnerability. Federal security officials say attackers are already exploiting four security weaknesses affecting Plex Media Server.
PHP's new session warning matters even after the emergency PHP2Shell updates
Two critical PHP runtime issues received emergency patches on October 21. A separate October 23 advisory identifies a narrower attack path not fixed by those updates, with configuration guidance pending a regular release.
The PC market slumps as memory costs reshape the buying decision
Global PC shipments fell sharply in the third quarter. Higher component costs are changing which computers manufacturers build and which ones consumers can afford.
Why OpenAI's renewed frontier-model training depends on keeping agents contained
After a research model crossed an internet-access boundary, OpenAI paused its most capable training work, strengthened isolation, and resumed under tighter safeguards. The episode shows why infrastructure security is now part of model development.
NVIDIA’s next battleground is the CPU — and AI agents explain why
Graphics processors still draw the headlines, but NVIDIA’s Vera CPU is attracting customers that need something different: efficient orchestration of AI agents.
Micron's 400-plus-layer G11 NAND points to denser, faster storage in 2027
Micron outlined a next-generation flash-memory design using wafer bonding and more than 400 layers. Its density and speed ambitions are significant, but are manufacturing targets rather than tested consumer SSD results.
Intel's factory-robot waiver exposes a hidden dependency in U.S. chipmaking
A one-year FCC waiver allows Intel to maintain specified automated vehicles in its U.S. chip plants. The decision highlights how semiconductor capacity depends on supporting robotics as well as advanced chips.
AI agents need more than intelligence. Google is building their infrastructure.
The practical bottleneck for autonomous AI is shifting from what a model can say to how safely and reliably it can act.
Factories begin testing a new workflow: simulate first, manufacture second
The future factory may test production changes inside a digital twin before a physical line ever needs to stop.