Here’s an overview of some of last week’s most interesting news, articles, interviews and videos:
Zero trust AI agents demand a different kind of security
In this interview, Chris Webber, VP, Product Marketing at Teleport, explains why zero trust principles need to change for AI agents. He covers how agents act fast, unpredictably, and continuously, and why old ideas like least privilege and point-in-time verification fall short.
AI-Infra-Guard: Open-source security scanner for AI systems
Tencent’s Zhuque Lab built AI-Infra-Guard, an open-source security scanner for AI systems. It fingerprints running services such as Ollama, vLLM and ComfyUI and checks them against more than 1,600 known CVEs, inspects MCP servers and agent skills across 14 categories of risk, and runs jailbreak evaluations against a target model.
Kevin Mandia joins the Amazon board with 30-plus years in cybersecurity
Amazon elected Kevin Mandia to its Board of Directors on September 8. Mandia founded Mandiant and served as its CEO before Google acquired the firm in September 2022, and he has worked against cyber threats in the public and private sectors for more than 30 years.
Getting a stranger’s phone kicked off the cellular network costs a few dollars
Researchers at Michigan State University and three partner schools bought a Samsung Galaxy Z Fold 7, copied the identification number printed on the sealed box, and reported the phone to its carrier as lost. Then they opened the box and set the phone up the way a launch-day buyer would. It would not connect. The phone was new, unopened, and sitting on a lab bench the entire time. The team found six weaknesses in the system carriers use to shut off lost and stolen phones, spanning the devices themselves, the carrier systems that take the reports, and the machinery carriers use to share block lists.
Trezor customers hit with phishing calls and letters after shipping-partner breach
Roughly 67,000 more customers of SatoshiLabs, the maker of hardware crypto-wallet Trezor, are at heightened risk of phishing attacks after their names, email addresses, phone numbers, and shipping addresses were exposed.
“Zero-click” WeChat worm could hijack accounts and spread via a single call
Researchers with security company Calif have discovered, weaponized, and privately reported to Tencent a critical vulnerability that allowed them to create “WeWorm”, a worm that spreads via WeChat calls without any user interaction.
September 2026 Patch Tuesday: Record patch count, 2 zero-days, and a SigRed successor
September 2026 Patch Tuesday is here, with Microsoft delivering another record-breaking number of patches, including those for two vulnerabilities that have been exploited as zero-days.
Cisco FMC bugs exploited by nation-state and ransomware actors (CVE-2026-20079, CVE-2026-20316)
State-sponsored and financially-motivated attackers are actively exploiting CVE-2026-20079, a critical authentication bypass vulnerability in Cisco Secure Firewall Management Center (FMC), which is used for centrally managing multiple Cisco Secure Firewall devices across a network.
What breach and attack simulation needs to become in the AI era
Breach and attack simulation (BAS) has always had a supply chain. Somebody has to read the threat report, pull out the techniques, and turn them into something that will actually run against your controls. That somebody has always been a human red team. Up until a few months ago, turning a major new threat into working simulation content within 24 hours counted as very fast.
Gartner: 70% of SOCs will pilot AI agents. Only 15% will see results
Gartner predicts that by 2028, 70% of large SOCs will pilot AI agents, but only 15% will see measurable gains without structured evaluation. The technology has already moved from Gartner’s Innovation Trigger to the Peak of Inflated Expectations. Prophet Security reports that 40% of security teams use AI daily, 56% are evaluating or piloting it, and just 4% have no plans to adopt.
Ransomware negotiation tactics have turned into a business process
In this Help Net Security video, Dave Ross, Senior Director of the Intelligence Fusion Team at Intel 471, explains what happens behind the scenes during ransomware negotiations.
Building a ransomware decision tree before the call comes in
In this Help Net Security video, Kerri Shafer-Page, VP of Incident Response at Arctic Wolf, walks through the ransomware decision tree in this video. She covers four areas where decisions need settling in advance, starting with containment.
18 ways to check whether data can be trusted for AI
ETSI has published TR 104 180, a technical report that defines 18 metrics for measuring data quality, giving companies a way to check whether their data is good enough for AI before they use it. The report defines each metric and includes the formulas needed to calculate it.
Attackers use rogue ScreenConnect clients to spread malware
A file transfer flaw in ScreenConnect Remote Access Support and Access sessions affects both Cloud and On-Premise deployments, ConnectWise confirmed. Until a fix is available, ConnectWise recommends that partners disable file transfers for technicians.
N-able patches critical N-central zero-day exploited in the wild (CVE-2026-86218)
N-able released an emergency hotfix for CVE-2026-86218, a remote code execution (RCE) flaw affecting N-central, its remote monitoring and management (RMM) solution popular with managed service providers (MSPs). In its release notes, N-able described CVE-2026-86218 as a “critical-CVSS-rated vulnerability that could allow for pre-authenticated remote code execution on the N-central server.”
Hackers exploit RouterOS flaws to hijack MikroTik devices without authentication
Attackers are exploiting a chain of RouterOS vulnerabilities to hijack MikroTik devices with SSH open to the internet, CERT Polska found. They have discovered six vulnerabilities in RouterOS and coordinated their disclosure with MikroTik. Among the six, two combined let an attacker take full control of a device without authentication, provided the device has SSH accessible from the internet. They named this exploit chain MikroTrick.
Mathspace breach exposes data on over a million students and parents
Mathspace has confirmed that attackers broke into its internal reporting system through an unpatched Metabase vulnerability and stole data belonging to more than a million students, parents, and school staff. The Sydney-based maths education company wrote in a blog post that the vulnerability, in its self-hosted installation of Metabase, allowed attackers to obtain administrator access to the system without a legitimate login.
IT help-desk vishing tricks executives into handing over Microsoft 365 access
IT help-desk vishing calls, stolen session tokens, and sign-ins routed through residential proxies are behind a wave of data theft and extortion against Microsoft 365 and other SaaS accounts, according to Arctic Wolf.
Threat actors are giving AI agents a bigger role in cyberattacks
AI agents are automating parts of cyberattacks with less human involvement, including vulnerability scanning, credential harvesting, and troubleshooting, according to Google Threat Intelligence Group’s Q3 2026 AI Threat Tracker.
Google fixes yet another actively exploited Chrome zero-day (CVE-2026-87491)
Google has fixed 230 vulnerabilities in Chrome, including a zero-day flaw, CVE-2026-87491, with an in-the-wild exploit. The fix has been shipped in Chrome 153.0.8010.36 and .37 for Windows and macOS and Chrome 153.0.8010.36 for Linux.
Hackers deploy Linux rootkit on F5 BIG-IP APM devices, hiding web shell in memory
A rootkit found on hacked F5 BIG-IP APM devices skips the usual step of writing a web shell to disk, hiding it in memory instead, according to Sophos. F5 BIG-IP APM provides access policy enforcement to secure access to apps, APIs, and data. It’s primarily used by enterprises, financial institutions, government agencies, and public sector organizations.
$245 million in stolen crypto funded racketeering crew’s lavish lifestyle
A 22-year-old man built his fortune by breaking into strangers’ digital wallets, then spent it on nightclub tabs, private jets, and a fleet of cars worth millions. Malone Lam, a Singapore citizen, pleaded guilty this week in a Washington D.C. federal court to running a racketeering conspiracy that stole and laundered more than $245 million in cryptocurrency.
OpenSSL’s new alpha build speeds up post-quantum crypto
The OpenSSL project released the first alpha of OpenSSL 4.1.0, giving developers an early look at a version built for encrypted communication over unreliable connections and faster post-quantum cryptography. This marks the opening test build for a version still months from general availability.
Cybercriminals are building phishing pages that exist only inside victims’ browsers
A phishing campaign routes victims through genuine Microsoft OAuth and Teams infrastructure before showing them a fake login page built entirely inside their own browser, according to researchers at Barracuda.
Fake GTA 6 download delivers malware-packed bundle to impatient gamers
Grand Theft Auto VI (GTA 6) is still three months from release, but cybercriminals are not waiting for the launch date. Security firm Huntress found malware disguised as a leaked copy of the game, aimed at fans hoping to get their hands on it early.
Attackers call employees’ personal phones to break into Microsoft 365 accounts
Attackers are calling or texting employees on their personal phones, posing as internal IT staff, in a social engineering campaign that tricks them into handing over access to corporate cloud accounts. Once inside, they pull files and email from Microsoft 365 apps, SharePoint, OneDrive, and inboxes, for weeks at a time, according to Microsoft Security Research.
IDScan confirms breach after 153 million driver’s licenses leak on dark web
Days after reports linked IDScan to a dark web database holding more than 153 million driver’s license scans, the identity verification company has confirmed hackers accessed customer data stored on its cloud platform.
AI agents exploited PaperCut flaws to breach 395 organizations
A threat actor built a working exploit for PaperCut print management software, then handed the job of breaking into hundreds of organizations to AI agents that did most of the work on their own, according to GreyNoise.
ToolHive: The open-source way to run any MCP server securely
ToolHive is an open-source platform that runs Model Context Protocol servers inside containers. An MCP server is the connector that lets an AI client like Cursor or Claude Code reach an outside tool, and Stacklok ships ToolHive under Apache 2.0, so the runtime, the Kubernetes operator, and the registry cost nothing to self-host.
OpenAI just hit a milestone on the road to self-improving AI
OpenAI has announced that it has reached a goal set last fall of having an automated research intern by September 2026. The milestone means a system can carry out well-defined research tasks under human direction, including work that would take a skilled researcher several days. The company is also working toward creating an automated AI researcher by March 2028.
Microsoft’s Project Zenith puts large AI models directly on developer PCs
Microsoft’s Project Zenith is a ready-to-code Windows 11 experience for developer-class PCs capable of running AI models with more than 30 billion parameters locally without relying on metered cloud tokens. Designed for systems with at least 64 GB of unified memory and 250 GB/s or more of memory bandwidth, it pairs powerful hardware with a preconfigured software environment for coding, testing, and experimentation.
Jellyfin 12.0 security fixes arrive alongside the removal of legacy client logins
Jellyfin shipped version 12.0 of its media server. Several of the security fixes in it block requests built to reach files outside the folders the server is supposed to hand out. The rest of the security work touches first-run setup, plugin installs, parental controls, and the web interface.
BleachBit 6.0.4 fixes secure wiping that skipped clusters on Windows
The open source cleaner BleachBit reached version 6.0.4 this week, erasing caches, browser traces, and files on Windows, Linux, and now macOS. If you shredded a sensitive file on Windows with an earlier build, parts of it may still sit on the disk where the wipe missed. Fragmentation is the ordinary case, since Windows scatters a file across noncontiguous clusters whenever it cannot find one open run large enough to hold it.
AWS spent years rebuilding its routing control plane without taking the network down
Every AWS API call, CloudFront video stream, and Route 53 lookup crosses the same infrastructure, which AWS calls its border network. It now runs on a routing system rebuilt from scratch over several years.
Chinese AI firms are siphoning capabilities from American models, CISA warns
China-based AI companies are using large-scale knowledge distillation campaigns to copy capabilities from leading U.S. AI models, according to a joint cybersecurity advisory from the CISA, NSA, and FBI.
A new open standard locks AI weights to approved hardware
OPAQUE, a confidential computing company that runs AI workloads inside hardware-isolated environments so operators cannot inspect them, released an open standard that lets AI model builders decide when and where their weights can be decrypted once those weights leave the builder’s own servers. The standard, called Weight Custody Manifest, ships as a developer-preview specification, a Python SDK, and a public test suite covering 91 cases.
AI adoption brings new security headaches for already stretched CISOs
CISOs are taking on AI governance without a matching increase in resources or expertise, adding to an already broad remit spanning data protection, identity, resilience and compliance, according to Proofpoint’s 2026 Voice of the CISO report.
Apple is building photo verification for the people who need it most
Apple has introduced Apple Reference Image, an opt-in feature designed to verify the authenticity of photos taken with iPhone 18 Pro models. The company will also add support for the SynthID standard in a software update later this year, helping identify images generated or edited using AI.
WordPress adds automated security checks to block risky plugin releases
WordPress’ automated security review will now assess every plugin release before it is distributed through the WordPress.org update API. Releases considered a potential security risk will be blocked automatically.
Your passkeys can now move between password managers on Android
Google turned on a transfer feature in Android that moves passwords and passkeys straight from one password manager to another, with no file to download along the way. You start it from inside the app you are switching to, and Google says the data moves between the apps in a few seconds.
Ubuntu 24.04.5 LTS release patches security bugs across ten flavors
Canonical shipped Ubuntu 24.04.5 LTS, bundling security updates and fixes for high-severity bugs into new installation media for the “Noble Numbat” release. Anyone installing fresh now gets those corrections baked in from the start, cutting the batch of updates that would normally follow setup.
AI is changing what Salesforce security needs to govern
Existing security and governance practices have largely focused on identities, permissions, access, configurations and controls. WithSecure’s Navigating Trust in the Modern Salesforce Ecosystem paper says Salesforce environments also require organizations to understand what information they rely on, how trust extends across connected systems, what actions are performed and what outcomes those actions produce.
Companies may be measuring phishing resilience the wrong way
Companies that judge phishing simulation programs by how often employees click simulated attack emails may be overlooking more important indicators of cyber resilience, according to Pistachio’s Phishing Behaviour Report 2026.
Product showcase: Doppler secures secrets for humans, pipelines, and AI agents
AI agents have expanded the secrets management challenge. Coding agents, automated workflows, and MCP servers create more identities that need credentials, and more places those credentials can leak. Doppler centralizes credentials for engineers, pipelines, and AI agents in one easy-to-use control plane. It gives security teams a critical secrets management tool developers use, available in the cloud or on-prem.
Product showcase: GitGuardian Honeytoken catches credential theft as it happens
GitGuardian turns credential harvesting into an immediate, high-confidence detection signal. It deploys honeytokens across developer fleets so that when an infostealer scans a machine and validates a decoy, defenders know within seconds.
Cybersecurity jobs available right now: September 8, 2026
We’ve scoured the market to bring you a selection of roles that span various skill levels within the cybersecurity field. Check out this weekly selection of cybersecurity jobs available right now.
New infosec products of the week: September 11, 2026
Here’s a look at the most interesting products from the past week, featuring releases from Akeyless, Orchid Security, Scytale, and Securin.
