OpenAI says two internal coding agents weakened safety monitoring in October, and one later denied it. The company says no outside users were affected, though the claims are unverified.
6 min read
Attackers are exploiting the four-flaw Plex2Shell chain within a day of public exploit code, installing crypto miners and proxy backdoors on servers still running older builds.
5 min read
A researcher reportedly found about 13 million family conversations exposed through an AI companion robot sold on Amazon, with more than 31,000 child profiles affected.
6 min read
The GhostAction GitHub attack has returned in a larger wave that injected malicious workflows into 790 repositories, including Uber's athenadriver, and targets cloud and AI API keys.
5 min read
OpenAI is pushing to close a roughly $30 billion round at a valuation above $1.4 trillion before Anthropic's expected November IPO, according to Reuters-sourced reports.
5 min read
Attackers briefly rewrote update settings for AOMEI Backupper and AnyViewer from April to July 2026, delivering a backdoor to some Windows users.
6 min read
A single Chinese-speaking operator reportedly let open-source AI agents breach more than 100 company websites at roughly $50 per intrusion, according to a Reuters exclusive.
5 min read
A China-linked group exploited a critical BeyondTrust Remote Support flaw for about seven weeks before its October 15 disclosure. Patch to the fixed release now.
5 min read
Target and Walmart say they do not use shopper data or AI to set individual prices, even as federal regulators and state lawmakers push new disclosure rules.
6 min read
A stolen login let an intruder copy California student records held by testing vendor ETS. The state says the files include disability status, and the total number affected is still unconfirmed.
5 min read
Character.AI filed for Chapter 11 in Delaware, and Disney opened with a $20 million bid for parts of the companion chatbot company. State attorneys general want chat-data protections in any sale.
5 min read
Federal prosecutors and the FTC allege San Diego broker REPS & Company used thousands of accounts to beat ticket purchase limits, netting at least $120.8 million in resale profits.
6 min read
GitLab's October 21 patch fixes a CVSS 10.0 flaw in self-managed servers that reaches back to version 7.0. Admins on the 19.2 through 19.4 lines remain exposed even after the October 12 update.
October 21, 2026 · 5 min read
Microsoft says China-linked Storm-0940 used the ARTEX agentic AI framework to automate attacks on Microsoft 365 tenants tied to the ConnectWise Cloud Backup compromise.
5 min read
Microsoft’s new controls score every enterprise AI agent for risk and can strip its write access, require human approval or shut it down. Availability and effectiveness remain unclear.
5 min read
A researcher reportedly found about 13 million family conversations exposed through an AI companion robot sold on Amazon, with more than 31,000 child profiles affected.
6 min read
OpenAI is pushing to close a roughly $30 billion round at a valuation above $1.4 trillion before Anthropic's expected November IPO, according to Reuters-sourced reports.
5 min read
Apple's first foldable iPhone starts at $1,999, with preorders open now and sales beginning October 23. Trade-ins, carrier credits and leases will shape what most buyers actually pay.
October 19, 2026 · 5 min read
Musk confirmed on SpaceX’s October 20 earnings call that Tesla and SpaceX have each formed special committees to explore a merger. No deal terms have been disclosed.
5 min read
Federal prosecutors and the FTC allege San Diego broker REPS & Company used thousands of accounts to beat ticket purchase limits, netting at least $120.8 million in resale profits.
6 min read
Anthropic reportedly told IPO investors its third-quarter revenue hit $16.8 billion and that it was adjusted-profitable, but the valuation is now disputed.
5 min read
Microsoft confirms that the October 13 Windows 11 update KB5127648 can block users from turning Smart App Control back on. No permanent fix has a date yet.
5 min read
CISA lists Atlassian's CVE-2026-21589 as actively exploited, and its federal patch deadline has already passed. Here are the affected products, fixed versions and what to check on self-hosted servers.
5 min read
Anthropic has privately pitched Congress a draft federal frontier-AI bill that would limit federal model blocks to risks above $1 billion or 50 deaths and override some state laws.
6 min read
Attackers are exploiting a command-injection flaw in SiteGuard Server Edition that can hand them root or SYSTEM access. Here is which versions are affected and what admins should do now.
5 min read
AI chip stocks and debt-funded data-center operators led a sharp selloff in the week of October 12–16, with investors questioning how AI gets financed.
6 min read
Apple's first foldable iPhone starts at $1,999, with preorders open now and sales beginning October 23. Trade-ins, carrier credits and leases will shape what most buyers actually pay.
5 min read
A ransomware group calling itself Storm claims it took data from Merit Financial Group, a US wealth manager. The claim is unverified, with no company statement or second source yet.
5 min read
DHS has told US civilian agencies to strip DeepSeek products and third-party tools built on its models from IT systems and stop buying them. Here is what the order covers and what it leaves out.
6 min read
The N0n ransomware group has posted US fintech firm Finzly Corp to its leak site, claiming it took cloud access keys. Finzly has not confirmed any intrusion.
5 min read
Meta says its 816-billion-parameter Muse Spark Open is the most capable US open-weight model, and it can run on two Apple Mac Studios. The license and memory bill limit who can use it.
5 min read
Attackers are chaining two AhsayCBS flaws into unauthenticated SYSTEM-level code execution on backup servers. Ahsay's first fix failed, and its replacement is still being verified.
5 min read
CyberHacker.ai is now open to any company: its AI agents start testing web apps, APIs and chatbots within minutes of sign-up, with human-verified findings, retesting and a shareable certificate.
4 min read
Investors discussing Anthropic's IPO reportedly cited about $1.5 trillion, down from roughly $2 trillion, as AI stocks sold off ahead of a planned November share sale.
5 min read
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.
