AI
Only 42% of Alpha School Students Hit Its 2x Learning Promise
A New York Times investigation found 42% of Alpha School students met its 2x learning promise, and critics dispute how the growth figure is calculated.
4 min read
Latest
USA 24: technology, cybersecurity and AI news
Top stories
More →
AI
Node2Shell: Patch Node.js Now for Brotli Flaw CVE-2026-111497
Public exploit code for a new Node.js remote-code-execution flaw appeared a day after the October 21 patch. Here is who is most exposed and which releases fix it.
5 min read
Tech News
Nvidia AI Financing Risk Grows as Neocloud Stocks and Bonds Slide
Nvidia has disclosed $36 billion in neocloud backstops and a $105 billion Ohio guarantee, and the CoreWeave debt selloff is now testing them.
7 min read
Cybersecurity
Navigate360 Breach Exposed School Crime-Tip Records Across US and Canada
A months-long breach at school-safety vendor Navigate360 exposed anonymous tip records in Nevada and beyond. About 10% of SafeVoice Nevada tips containing personal data were compromised.
5 min read
Latest news
More →
AI
OpenAI Will Stop Training Frontier Models on the Live Internet
OpenAI will train its most advanced models on cached and simulated copies of the web instead of the live internet, following a September 20 incident in which a model reached the open internet.
5 min read
Cybersecurity
China-Linked Storm-0940 Targeted 2,250 Microsoft 365 Tenants
Microsoft says China-linked Storm-0940 used a stolen ConnectWise app credential and an open-source AI hacking framework to probe more than 2,250 Microsoft 365 tenants.
6 min read
Cybersecurity
PwC Confirms Vendor Breach as Everest Claims 590 GB of Staff Data
Everest says it took about 590 GB of PwC HR and training data through a vendor. PwC confirmed a vendor incident but has not verified the claimed size of the theft.
5 min read
Innovations
Tesla's Robotaxi Safety Report Claims Zero Injuries in 3.1 Million Miles
Tesla says its robotaxis logged 3,112,541 driverless miles with zero injuries, but its benchmark covers only Texas and the report excludes miles with safety monitors.
6 min read
AI
Meta's $825 Billion AI Commitments Hit a Shaky CoreWeave Debt Market
Meta's roughly $825 billion in future leases and contracts, plus a CoreWeave bond selloff, pushed AI financing fears from niche neocloud stocks into the broader market.
6 min read
Cybersecurity
Stolen Microsoft 365 Tokens Breached 79 Organizations Without Passwords
A Huntress investigation found attackers reused stolen Microsoft 365 sign-in tokens to break into business email at 79 organizations between July and September, bypassing MFA and passwords.
5 min read
AI
Reservoir Sues Suno and Anthropic as Lyrics Fair-Use Ruling Looms
Music publisher Reservoir sued Suno and Anthropic on October 21, adding a new front to AI copyright fights while a judge weighs Anthropic's fair-use defense in a lyrics case.
6 min read
Cybersecurity
Oneonta City Schools Closed Until Tuesday After Ransomware Attack
A ransomware attack closed Oneonta, New York's schools on Friday, and students will not return until Tuesday, October 27. Officials have not said whether personal data was taken.
4 min read
Cybersecurity
More →ConnectWise Cloud Backup Breach Gave Attackers Global Admin Rights
Cloud Backup has been offline since October 17 after attackers abused a stolen Microsoft 365 API secret. ConnectWise is resetting credentials, but it has not set a restoration date.
October 25, 2026 · 6 min read
Artificial Intelligence (AI)
More →AI
DHS Orders Federal Agencies to Pull DeepSeek, Including Third-Party Tools
DHS has ordered US civilian agencies to stop buying DeepSeek products and remove them, including third-party tools built on its models. Private use is not covered.
6 min read
AI
Block's Dessa Is a Speech Model Trained Only on Conversation
Block says its Dessa model listens and talks at the same time, learned from human conversation alone, and runs locally on a high-end MacBook Pro. Every claim is company-stated for now.
5 min read
AI
Node2Shell: Patch Node.js Now for Brotli Flaw CVE-2026-111497
Public exploit code for a new Node.js remote-code-execution flaw appeared a day after the October 21 patch. Here is who is most exposed and which releases fix it.
5 min read
AI
OpenAI Will Stop Training Frontier Models on the Live Internet
OpenAI will train its most advanced models on cached and simulated copies of the web instead of the live internet, following a September 20 incident in which a model reached the open internet.
5 min read
Innovations
More →Tesla's Robotaxi Safety Report Claims Zero Injuries in 3.1 Million Miles
Tesla says its robotaxis logged 3,112,541 driverless miles with zero injuries, but its benchmark covers only Texas and the report excludes miles with safety monitors.
October 24, 2026 · 6 min read
Tech News & Analysis
More →Tech News
Nvidia AI Financing Risk Grows as Neocloud Stocks and Bonds Slide
Nvidia has disclosed $36 billion in neocloud backstops and a $105 billion Ohio guarantee, and the CoreWeave debt selloff is now testing them.
7 min read
Tech News
Meta Delays Commercial Terms After Publishers Object to AI Use
Meta is pushing its updated business terms to November 20 after publishers and advertisers said the rules would give it broad rights to use their content for AI.
6 min read
Tech News
SpaceX AI Revenue Passes Starlink, but Shares Give Back Earnings Gains
SpaceX's first earnings report as a public company showed its AI unit overtaking Starlink in revenue while absorbing most of its capital spending. Shares then gave back their gains.
6 min read
Tech News
Why Meta Stock Fell 7% Despite Record Revenue and a Stock Split
Meta's shares fell about 7% on October 22 despite a revenue beat, as $34.9 billion in quarterly capex pushed free cash flow negative for the first time since its 2012 listing.
5 min read
Trending topics
More stories
More →
AI
OpenAI Fired Three Safety Researchers Over Alleged Signal Chat Leaks
OpenAI fired three safety researchers on October 1. A Wall Street Journal report now says the company suspected them of leaking model-design details over disappearing Signal chats.
5 min read
Cybersecurity
Oracle's October Patch Fixes 1,209 CVEs, Including Six CVSS 10.0 Flaws
Oracle's October patch bundle covers 1,209 CVEs and six CVSS 10.0 flaws, including an E-Business Suite bug that needs no login. IT teams should start with the E-Business Suite and CVSS 10.0 items today.
4 min read
AI
American Express to Cut 5% of Its Workforce to Prepare for AI
American Express will cut about 5% of its global workforce and take a $336 million restructuring charge, with its CFO tying the cuts to preparing for AI.
5 min read
Cybersecurity
k0t Hackers Harvest 70,000 Credentials From FortiGate and GlobalProtect
Group-IB and NyxLab tied a suspected access broker called k0t to exposed FortiGate, FortiWeb and GlobalProtect devices. Rotate LDAP and admin credentials now.
5 min read
AI
Anthropic Junior Technical Hires Fell From 27% to 9%, Report Finds
Anthropic's internal study says junior staff fell from 27% to 9% of new technical hires as AI agents took on more work. The report is descriptive, not proof of cause.
5 min read
Cybersecurity
Trump Appointees Asked CISA to Secretly Hack a Voting Machine
DHS political appointees reportedly asked CISA to secretly buy and hack a voting machine, withholding flaws from the vendor. Career staff resisted, and the work never happened.
5 min read
AI
AI-Linked Layoffs Reach About 102,800 Jobs Across 31 Companies
Reuters' running factbox now counts about 102,800 announced job cuts at 31 companies since October 2025, with HSBC and Amazon leading the list.
5 min read
Tech News
Meta Delays Commercial Terms After Publishers Object to AI Use
Meta is pushing its updated business terms to November 20 after publishers and advertisers said the rules would give it broad rights to use their content for AI.
6 min read
Cybersecurity
Nextcloud Collabora Flaw Under Attack: Patch CVE-2026-110447 Now
Attackers are planting PHP web shells through the built-in CODE app in Nextcloud. Update to 26.4.303 or remove the app, and check your server for signs of compromise.
5 min read
AI
Texas Sues OpenAI and Sam Altman, Joining Eight States Over ChatGPT
Paxton says OpenAI and Sam Altman misled Texans about ChatGPT's risks, adding a state to the child-safety suits that eight other states filed on October 14.
5 min read
Cybersecurity
Cisco Secure Email Gateway Zero-Day Under Attack: Patch CVE-2026-76519
Cisco says attackers are exploiting CVE-2026-76519, a CVSS 9.8 flaw that gives an unauthenticated attacker root through a booby-trapped email attachment. There is no workaround, so patching is the only fix.
5 min read
AI
AI Found Two PHP Flaws, and NSA and FBI Warn Sites to Patch Now
AI agents helped find two critical PHP flaws that could give attackers access to server files and databases. The NSA and FBI say take affected sites offline if they cannot be patched right away.
6 min read
Technology news that respects your time
Every USA 24 story is built on cited sources and written to answer the questions that matter: what happened, why it matters, and what you should do next. We cover the stories shaping how Americans work, stay safe online and spend their money on technology.
Sourced
Every article links the documents and reporting it relies on.
Practical
Clear steps when there is something you should do.
Updated
Stories are corrected and updated as facts change.