Technology

AI Story Leads Collapse Under Basic Checks, Highlighting Hallucination Risk in Newsrooms

An in-house AI tool produced six detailed but nonexistent news leads. Editors traced them to hallucinations, underscoring the need for strict verification.

#ai, #journalism, #misinformation, #verification

Back-to-back Starlink launches push SpaceX past 11,000 satellites, deepening debates over security and sky crowding

Two Falcon 9 missions in 13 hours grew Starlink’s fleet past 9,600 active satellites, underscoring its role in war, broadband and orbit policy.

#spacex, #starlink, #satellites, #spacepolicy, #nationalsecurity

International AI safety report warns frontier systems are outpacing oversight

A major international report says frontier AI is advancing faster than testing and governance, raising risks from deepfakes to cyber and bio misuse.

#ai, #aisafety, #deepfakes, #cybersecurity, #governance

Meta’s $135 Billion AI Spending Plan Raises the Stakes for Big Tech—and the Power Grid

Meta projects up to $135B in 2026 capex for AI data centers, jolting rivals and suppliers while intensifying pressure on chips, energy and regulation.

#meta, #artificialintelligence, #datacenters, #capex, #powergrid

Google Tests Gemini-Powered ‘Auto Browse’ Agent Inside Chrome

Google is rolling out an experimental Chrome feature that lets Gemini click, scroll and fill forms to complete online tasks—raising new security and privacy questions.

#google, #chrome, #gemini, #aiagents, #cybersecurity

Nvidia opens ‘Earth-2’ AI weather models, promising faster forecasts on smaller GPU clusters

Unveiled at the AMS meeting in Houston, Nvidia’s open Earth-2 models aim to speed forecasting and cut costs—raising new reliance questions.

#weather, #ai, #nvidia, #forecasting, #climate

EU targets Google’s Android AI and search data in new Digital Markets Act proceedings

EU regulators launch proceedings to force Google to open Android AI access points and share anonymized search data, tightening DMA enforcement.

#eu, #google, #dma, #android, #ai

Waymo Robotaxi Hits Child Near Santa Monica School, Prompting Federal Safety Probe

A low-speed collision near an elementary school has prompted NHTSA to investigate how Waymo’s driverless cars handle school zones and children.

#waymo, #robotaxi, #selfdriving, #nhtsa, #schoolsafety

Fortinet disables FortiCloud SSO worldwide after critical cross-tenant login flaw is exploited

A critical FortiCloud SSO bug let attackers cross into other customers’ Fortinet devices. Fortinet flipped off SSO globally, then rushed patches.

#cybersecurity, #fortinet, #vulnerability, #sso, #cisa

A neuroscientist wants to train AI on brainwaves—and privacy law may be the biggest obstacle

A Swiss researcher proposes training foundation models on fMRI and EEG signals to align AI with human reactions, raising new stakes for neural privacy.

#artificialintelligence, #neuroscience, #privacy, #neurotechnology, #regulation

When AI Pitches the News: A Politics Desk Spikes Six Plausible Stories That Never Happened

Editors killed six AI-assisted politics leads after basic checks found no documents or corroboration—highlighting the newsroom’s verification line.

#ai, #journalism, #factchecking, #disinformation, #unitednations

Intel Ships First U.S.-Made 18A Laptop Chips as ‘AI PC’ Push Hits Stores

Intel’s Panther Lake Core Ultra Series 3 debuts Jan. 27, bringing 18A chips made in Arizona—and Copilot+ ready AI—to premium laptops.

#intel, #aipc, #chipsact, #semiconductors, #laptops

Newsrooms Turn to AI That Knows When to Say ‘I Don’t Know’

Publishers testing generative AI are finding the safest tools are those that flag missing facts, disclose limits and refuse to fabricate news copy.

#journalism, #generativeai, #newstrooms, #aistandards, #mediaethics

Blue Origin’s New Shepard flies six to the edge of space, bringing total to 98 people

New Shepard’s NS-38 carried six passengers above the Kármán line from West Texas, adding to Blue Origin’s safety record and nearing 100 people flown.

#blueorigin, #new_shepard, #spaceflight, #westtexas, #jeffbezos

Rocket Lab launches Open Cosmos satellites, kicking off bid for ‘sovereign’ European broadband

Rocket Lab’s Electron placed two Open Cosmos comms satellites into a 1,050-km polar orbit, advancing a Ka-band plan for European ‘sovereign’ connectivity.

#space, #satellites, #rocketlab, #opencosmos, #broadband