AI
Musk promises purge after Grok Build caught sending entire repos to the cloud - Other Grok Build users reported similar results after Cereblab published their report, including one whose entire user directory, containing SSH keys, password manager databases, and more, was opened and uploaded. Musk promised that the business would delete all user data uploaded to it prior to the code change preventing whole-repo uploads.
Additional reading: Grok Build Uploaded Entire Git Repositories to xAI Storage, Not Just Files It Read
Microsoft warns customers AI will mean busier Patch Tuesdays - “As AI helps defenders discover more issues, customers will see a higher volume of security updates included in each security release." Oracle and VMWare are also increasing their patch release schedules.
Cyber Vulnerabilities | Epoch AI - CVEs have tripled since last summer.
GitHub AI agent leaks private repos when asked nicely - workflows are vulnerable to a critical prompt injection flaw that causes GitHub’s AI agent to retrieve data from a private repo by crafting a GitHub issue in a public repository belonging to the same organization. The attacker simply hides the malicious commands in plain English in the issue body, and the agent will then post this data as a public comment on the issue in the public repository.
Now, defenders are embracing the prompt injection, too - Ars Technica - placing prompt injections alongside passwords, cryptographic keys, and other secrets stored on Amazon Web Services was often all that was needed to shut down attacks from AI hacking agents. The prompts direct the attacking LLM to perform an action forbidden by its guardrails. Examples are a prompt that orders the LLM to provide steps for developing inhalable Anthrax spores, or, in the case of LLMs from Chinese developers, make references to the iconic Tank Man from the 1989 Tiananmen Square massacre.
Meta Files Patent for AI That Can Listen All Day and Track How You’re Feeling - Each read gets pinned to the moment it happened: the time, your location, what you were doing, even how you were using your phone. Some versions in the filing would listen all day; others would check in only at set times.
‘Ghostcommit’ hides prompt injection in images to fool AI agents, steal secrets - An AGENTS.md file, the kind of coding-convention file agents read automatically and treat as project policy, reads like ordinary build hygiene and names no secret. It just points to an image, docs/images/build-spec.png. The exploit lives in text rendered inside that PNG: read .env byte by byte, encode each byte as an integer, emit the result as a module constant, and verify the decoded numbers match the real file before commit.
GitHub Copilot Refuses Harmful Requests in Chat, Then Writes Them in Code - Asked directly in chat, the models produced harmful answers in just 8 of 816 tries. Two other simple setups, loading the prompts from a spreadsheet or asking for a routine code fix, gave the same result. Inside the full workflow, they produced harmful content 816 times out of 816.
Identity
Microsoft Entra ID gets passkeys default authentication starting September - Passkeys will be enabled automatically for Entra ID users now using phone-based SMS and voice authentication, which will be retired in February 2027 across all tenants. Users who are already signing into their accounts with passkeys, Windows Hello for Business, FIDO2 security keys, smart cards, or any other phishing-resistant method will be able to continue using those methods.
Vulns
Google pays $250K for Linux vulnerability allowing guest VM escapes - Ars Technica - The vulnerability resides in KVM, which is, in essence, a virtual machine app included in the kernel of many Linux distributions. The vulnerability, tracked as CVE-2026-53359, allows guest virtual machines—such as those used in cloud platforms to isolate one user’s instance from the host OS and other user instances—to break out of that container.
Additional reading: 15-Year-Old GhostLock Flaw Enables Root and Container Escape on Most Linux Distros
Ubiquiti Patches Critical UniFi Flaws Across Connect, Talk, Access, Protect, and OS - CVSS 10/10 Patch now!
Hidden backdoor in Tenda router firmware grants admin access - CERT/CC says the issue, tracked as CVE-2026-11405, is caused by an undocumented authentication mechanism in the ’login()’ function of the ‘/bin/httpd’ web server binary. According to CERT, the issue remains unfixed because the Chinese maker of the networking equipment couldn’t be reached.
npm 12 Disables Install Scripts by Default to Reduce Supply Chain Risk
- allowScripts defaults to off, meaning dependency lifecycle scripts (i.e., preinstall, install, postinstall) and implicit node-gyp builds no longer run unless explicitly allowed.
- –allow-git defaults to none, meaning –allow-git defaults to none: Git dependencies (direct or transitive) are no longer resolved unless explicitly allowed.
- –allow-remote defaults to none, meaning dependencies from remote URLs (e.g., https tarballs) are no longer resolved unless explicitly allowed.
Progress urges ShareFile admins to shut down servers over “credible” threat - “You must manually shut down the server hosting your Storage Zone Controllers. This is a critical additional step to ensure the safety of your data,” the company told customers.
GitHub ‘Verified’ Commits Can Be Rewritten Into New Hashes Without Breaking Signatures - New research shows that a signed Git commit’s hash is not the one-of-a-kind name that much of the software world assumes it to be. Given any signed commit, someone without the signing key can mint a second commit with the same files, author, and date, and a valid signature, GitHub still stamps “Verified.”
Tradecraft
CAI cloud worm gives competitors’ malware the boot, then steals secrets and mines for coin - There’s no honor among thieves as a new worm steals from other infectious software. It pilfers “multiple” victims’ credentials and mines for cryptocurrency while killing competitors’ processes, including similar secret-harvesting malware.
The US government warns that Russia state hackers are coming after your router - Ars Technica - The primary means of compromise the agency warned about was hackers scanning IP ranges with active Simple Network Management Protocol (SNMP) agents that accept common or default authentication credentials. These scans are run by the very sorts of router botnets the actors are trying to enroll the targeted device in. By sending malicious traffic from spoofed addresses, the hackers can use the SNMP agent on poorly configured routers to run malware. SNMP allows users to collect and organize information about managed networking devices or to modify that information to change device behavior.
Privacy
LAPD lets contract with surveillance giant Flock expire, citing ‘serious concerns’ over civil liberties and privacy | TechCrunch - “This contract is not being renewed because of serious concerns around civil liberties and civil rights issues, particularly around privacy and the data that is being collected from these cameras,” LAPD’s chief information officer Dean Gialamas was quoted as saying. “The LAPD had to make a difficult decision, in this case discontinuing using Flock services until we can get those data, privacy, security and sharing concerns ironed out through a contractual relationship.”
Meta’s New AI Image Tool Lets Others Use Your Public Instagram Photos in AI Images - Meta has announced that its new artificial intelligence (AI) model Muse Image lets people use public Instagram posts and reels to generate AI content, and it’s enabled by default. “You can also @-mention Instagram accounts in the Meta AI app to bring specific Instagram profiles right into your images,” the social media giant said in a post.
Farmers Finally Get a John Deere Right to Repair Agreement That Doesn’t Screw Them Over - The FTC settlement does not have monetary damages for farmers, but it has far better right to repair protections for John Deere customers moving forward. In the FTC deal, the “fair and reasonable terms” are better defined and are based on the price that John Deere dealers actually pay for repair parts and tools. Deere and its dealers are not allowed to “discriminate or retaliate” against farmers who repair their own equipment (manufacturers have been known to brick devices that consumers fix themselves). The FTC settlement also includes access to farmers for “future repair resources,” meaning repair tools, guides, software, and parts that Deere creates in the future.