Mandiant AI Risk and Resilience Report 2026 (Google Cloud; wired by The Hacker News 16 September) case study: after compromising a SaaS provider, an attacker hijacked an active AI coding-assistant session on a developer workstation. The assistant recommended a poisoned external package; once accepted, the attacker used the session to install an infostealer via a poisoned PyPI package, harvest GitHub OAuth tokens, and deploy the self-propagating Shai-Hulud worm across about 100 internal repositories (secret theft and programmatic exfiltration). Distinct from earlier Keyv-linked npm worm / Mini Shai-Hulud supply-chain desk notes. Defenders: treat AI assistant tool-install prompts as high-risk; constrain package installs; rotate GitHub tokens; audit recent repo automation. Primary: Mandiant/Google Cloud report.
Mandiant AI Risk and Resilience Report 2026 (Google Cloud)
ai cloud identity
Incidents
Wed 16 Sep
Updated Sun 25 Oct
Premier Medical Group of the Hudson Valley P.C. published a Notice of Data Security Incident: after disruption of some IT systems, investigation found an unauthorized party accessed certain files on 14 June 2026; on 14 July 2026 PMG determined files may have included patient names, contact information, dates of birth, health insurance information, provider names, internal patient IDs, dates of service, medication information, and treatment/diagnostic information. Law enforcement notified; enhanced safeguards and staff training cited. The HHS breach portal lists Premier Medical Group of the Hudson Valley P.C. (NY, healthcare provider) with 282,075 individuals affected, submitted 21 August 2026, as a hacking/IT incident involving a network server. How the attack occurred not disclosed. Primary: company notice; HHS OCR breach portal.
Premier Medical Group — Notice of Data Security Incident
breaches healthcare
Incidents
Tue 15 Sep
Updated Mon 26 Oct
eSentire Threat Response Unit blog (published 15 September 2026, 7:00 am US Eastern; 7:00 pm Perth) details GhostCode, a novel OAuth 2.0 device-authorization phishing kit that hijacks Microsoft 365 accounts after the victim completes legitimate MFA on Microsoft's real sign-in page. Observed chain: Salesforce contact-form lure posing as a procurement officer for BJ's Wholesale Club, sales follow-up, NDA pretext, then a WeTransfer link to a password-gated HTML lure using three evasion layers (junk padding, character-level HTML comment injection and an AES-256-GCM-encrypted redirect URL); a Cloudflare Turnstile gate filters scanners before the device-code page. Kit requests a user_code using the Microsoft Authentication Broker application ID, presents a polished fake document portal, and captures tokens once the victim approves the attacker's device. eSentire describes Primary Refresh Token capture and, in one intrusion, nine successful API calls with the stolen token across a 78-second period and three devices registered within 78 seconds, with rotating residential proxy IP addresses matching the victim's geolocation. Distinct from password-stealing kits; MFA does not stop the flow because the victim authenticates Microsoft directly. Defenders: restrict or block device-code grant where unused, alert on Authentication Broker device registrations, treat unexpected WeTransfer/NDA procurement mail as high-risk, review Entra ID sign-in and device logs. Primary: eSentire TRU. Correction (26 Oct 2026): removed the Cyber Security News credit and link; every detail on this card is from eSentire's report, and the title no longer describes the 78-second window as an MFA bypass time. (source: eSentire)
eSentire TRU — GhostCode device-code phishing kit (15 Sep 2026)
tech identity cloud
Incidents
Tue 15 Sep
Updated Thu 22 Oct
Cyber Daily exclusive (15 September 2026) names Australian dealer-management software firm Auto-IT as the third-party IT supplier behind the recent Storm ransomware wave against Australian car dealerships and machinery suppliers. Auto-IT told Cyber Daily a small number of customer environments were affected via unauthorised use of a third-party remote monitoring and management (RMM) tool; customers were named on a dark-web listing site with claims of accessed business data. Company says the incident is contained, customer environments remain secure and fully operational, forensic specialists were engaged, and it is working with the Australian Cyber Security Centre and impacted customers. Cyber Daily links the wave (listings from about 18 August) to dealers including Westco Motors Cairns, Ramsey Bros, Penfold Motors, Sharp Motor Group, Agrimac, and Macquarrie — several of which already have desk cards noting an unnamed third-party supplier. Primary: Cyber Daily with Auto-IT quotes; related desk cards: penfold-motors-storm-20260914, macquarrie-storm-20260903. Correction (22 Oct 2026): the headline previously said Auto-IT confirmed Storm ransomware; Auto-IT's statement does not name Storm, and the Storm link is Cyber Daily's reporting. (source: https://www.cyberdaily.au/security/14188-exclusive-aussie-software-firm-auto-it-confirms-customers-compromised-by-storm-ransomware-attack)
Cyber Daily — Auto-IT confirms Storm / customer environments (15 Sep 2026)
australia
Incidents
Tue 15 Sep
Updated Sun 25 Oct
Developer Janis Elsts's incident notice for Admin Menu Editor Pro (published 14 September 2026, last updated 20 October) says a malicious admin-menu-editor-pro.zip was uploaded to adminmenueditor.com on 14 September 2026 and showed up as a version 2.35 update; it included a new file, includes/wp-user-consent.php, that installs a web shell. A clean 2.36 was released the same day, but later on 14 September 2.36 was also compromised; the developer says the attacker may have had root-level access to the server, which he took offline. He says anyone who installed 2.35 should treat the site as compromised and anyone who installed 2.36 as potentially compromised, particularly if downloaded after 19:00 UTC; 2.34 should be clean. He estimated that about 230 customers were affected in the initial compromise, often with multiple sites each, and identified about 300 more potentially affected customers from downloads in or near the window. Signs he lists from user reports: a new /wp-content/object-cache directory holding a hex-named web shell, wp_ocache or _wp_ocache_ options, a hidden user named wp_ followed by hex characters, hex-named MU-plugins and a _wp_cconsent_tick WP-Cron event. He advises restoring from a backup made before 14 September, or deleting the plugin and those artefacts, then changing all WordPress passwords, regenerating keys and salts and rotating other credentials stored on the site. According to BleepingComputer (15 September), Elsts said 2.35 was available from about 06:00 to 13:00 UTC, its code also created a hidden user account, the malicious build was installed on at least 1,500 sites, and the free Admin Menu Editor does not appear to be affected. Updated 25 Oct 2026: the developer says the update server has been rebuilt, downloads, licence activation and renewals are working again, and version 2.37.1 (based on the last known safe version, 2.34) adds optional checksum and signature verification for plugin updates. Primary: developer incident notice; wire: BleepingComputer. Correction (25 Oct 2026): the card now leads with the developer's incident notice, which says 2.36 was also compromised and gives about 230 customers in the initial compromise plus about 300 more potentially affected; the 06:00–13:00 UTC window, the 1,500-site count and the free-edition note are credited to BleepingComputer. (source: Admin Menu Editor Pro developer incident notice)
Admin Menu Editor Pro — Security Incident Affecting Customers 2026-09-14 (developer notice; last updated 20 Oct 2026)
breaches cloud
Incidents
Tue 15 Sep
Updated Mon 26 Oct
BleepingComputer (15 September 2026) reports Houston-based utility CenterPoint Energy confirmed in an SEC filing that an unauthorized third party obtained personal information for a portion of its customers through an external-facing system. CenterPoint says electric and gas services were not impacted and does not expect a material business effect; it activated IR, engaged third-party experts, hardened systems, and notified law enforcement/regulators. Class-action complaints filed in US federal courts allege the incident window was about 17 August–1 September 2026. Primary wire: BleepingComputer; company confirmation: SEC filing as cited there. UPDATE (14 Oct 2026, US time): KPRC 2 Houston reports that a CenterPoint breach report published on the Texas Attorney General's website on Monday (12 October, US time) lists 1,062,060 Texans as affected, gives names, addresses, Social Security information and driver's licence numbers as the information involved, and indicates affected people had not yet been notified when it was submitted. A CenterPoint spokesperson told KPRC the company is still working with third-party experts to assess the scope of affected customer data and identify which customers will be notified. (source: KPRC 2 Houston (14 Oct), citing the Texas Attorney General) Correction (26 Oct 2026): removed the threat actor's own claims about how many records it took and how, and its later data release; CenterPoint has not confirmed them and no outlet has independently verified them. (source: BleepingComputer) Correction (26 Oct 2026): removed every SecurityWeek credit, link and detail; the affected-Texans count, data types and notification status from the Texas Attorney General's breach database are now credited to KPRC 2 Houston (14 Oct), citing the Texas Attorney General. (source: KPRC 2 Houston (14 Oct), citing the Texas Attorney General)
BleepingComputer — CenterPoint Energy confirms customer data stolen (15 Sep 2026)
breaches ot ics
Incidents
Tue 15 Sep
Updated Mon 26 Oct
BleepingComputer (15 September 2026) reports five alleged leaders of the Black Axe cybercrime syndicate — Perry Osagiede, Franklyn Osagiede, Osariemen Clement, Collins Otughwor, and Musa Mudashiru — were extradited to the United States to face wire fraud and money-laundering charges. Prosecutors allege a South Africa-based internet fraud campaign (about 2011–2021; the defendants were arrested in South Africa in 2021 at the US request and extradited on 11 September 2026, per BleepingComputer) using romance and advance-fee scams, aliases, dating sites, and VoIP numbers to target US victims, including coercion via threats to publish sensitive photos. US Attorney’s Office for the District of New Jersey press release linked from the wire: “Five Prominent Black Axe Members Extradited for Conspiring to Engage in Internet Scams and Money Laundering.” Law-enforcement action / charging story; not a fresh organisational data-breach notice. Primary: DOJ USAO-NJ PR; wire: BleepingComputer. Correction (26 Oct 2026): the campaign was earlier described as Cape Town-based; BleepingComputer places it only in South Africa. (source: BleepingComputer)
DOJ USAO-NJ — Black Axe members extradited (linked 15 Sep 2026 wire)
breaches
Incidents
Mon 14 Sep
Updated Sun 25 Oct
Clover Health Investments Corp., a PPO and HMO organisation with a Medicare contract, says in a notification letter filed with South Carolina's Department of Consumer Affairs that on 4 July 2026 it determined it had experienced a cybersecurity incident involving unauthorised access to certain company applications and data. Its investigation found that between 3 and 4 July 2026 an unauthorised party gained access to credentials belonging to three non-managerial employees and used them to access company applications and data; Clover says it detected the incident on 4 July, ended the access within hours, engaged third-party cybersecurity experts and notified law enforcement. Certain data was accessed and acquired, and a review completed on 28 August 2026 identified personal information about the people being notified; the published letter template does not list the types of data. Clover says it is unaware of any attempted or actual misuse. The US Department of Health and Human Services breach portal lists Clover Health (Minnesota, health plan) with 138,677 individuals affected, submitted on 14 September 2026, as a hacking/IT incident involving a network server and other locations. Primary: Clover Health notification letter (South Carolina Department of Consumer Affairs); HHS OCR breach portal. Correction (25 Oct 2026): the card now rests on Clover's notification letter and the HHS portal; a social-engineering description, a list of data types and a note about attribution from a wire report we cannot verify have been removed. (source: HHS) Correction (25 Oct 2026): the card date has moved from 5 October to 14 September 2026, the breach report submission date the HHS portal lists; HHS gives a US date with no time. (source: HHS)
Clover Health — Notice of Data Security Event (consumer letter, South Carolina Department of Consumer Affairs, PDF)
breaches identity
Incidents
Mon 14 Sep
Updated Mon 26 Oct
St James’ Anglican School (Alkimos, Western Australia) published an incident update dated Monday 14 September 2026 (9:00am): it identified a cyber security incident involving unauthorised access to its computer systems, contained and secured systems, and is auditing. Personal information relating to members of the school community was involved; families were notified with protective steps. The school says it reported to ASD’s ACSC and the OAIC; investigation ongoing; no technical commentary. The West Australian (15 September) reported that student medical records, photos and bank details were stolen; Insurance Business (17 September), citing The West Australian, says the data belongs to current and former students dating back to 2015 and includes names, addresses, email addresses, phone numbers, bank account details, student medical records and photographs. The school’s public statement does not itemise the data. The school’s statement does not name a threat actor. UPDATE 1 Oct 2026: according to Cyber Daily (David Hollingworth), the ThreeAM ransomware group claimed responsibility for the attack in a 28 September post to its darknet leak site; a school spokesperson told Cyber Daily experts have not identified evidence that any school information has been published or released online, investigation and dark-web monitoring continue, learning undisrupted, ACSC and OAIC informed at first discovery. Primary: school incident-updates page; wire: Cyber Daily 1 Oct; The West Australian 15 Sep; Insurance Business 17 Sep. Correction (25 Oct 2026): removed a quote from the ransomware group and a leak-site tracker reference; the group’s claim is now credited to Cyber Daily. (source: Cyber Daily) Correction (26 Oct 2026): the data categories are now attributed to The West Australian and Insurance Business, which are linked; a 'NewsWire' and DataBreaches.net credit and a breach-list reference that no linked source supported have been removed. (source: Insurance Business)
St James’ Anglican School — Incident Update (14 Sep 2026)
australia
Incidents
Mon 14 Sep
Updated Mon 26 Oct
Spain’s Agencia Española de Protección de Datos (AEPD) blog (14 September 2026; Francisco Pérez Bes) reports the agency’s first notification of a personal-data breach in which the incident was allegedly executed by an AI agent using a known large language model. Per the notifier: the agent searched generic files for vulnerabilities, successfully logged in, then autonomously hunted application flaws, modified personal data, and accessed invoices. AEPD stresses the claim is from the organisation’s notification and still requires analysis; use of a given model does not imply the provider’s model or infrastructure was compromised or purpose-built for crime. Framing: shift from AI-assisted attacker tools (phishing copy, vuln search) toward agentic chaining of attack phases at machine speed — with implications for identity/credential controls, detection, and response timing. National Cryptologic Center (CCN) paradigm-shift context noted in wire coverage. Primary: AEPD blog. Updated 17 Sep 2026: BleepingComputer reported the AEPD post on 16 September US time (17 September 1:26 am Perth). (source: BleepingComputer)
AEPD — primera notificación brecha por agente de IA (14 Sep 2026)
ai identity
Incidents
Mon 14 Sep
Updated Sat 24 Oct
Brevo's status write-up says that on 14 September 2026 an attacker used a compromised long-lived Cloudflare API key (stored in Brevo application source code; Brevo says it was first misused in late August) to deploy a Cloudflare Worker that rewrote responses at the CDN edge. Brevo gives the impact window as 15:01 to 20:30 UTC (5h29): the Worker was routed to all of brevo.com at 15:01, and at 16:07 it was updated to also append a loader to the three embedded JavaScript files and routed to sibforms.com; Brevo says the ClickFix was active on the listed URLs from 16:07 until 20:30 UTC. Affected: brevo.com, sendinblue.com, login/account/my/onboarding.brevo.com, sibforms.com, plus Brevo forms script, Conversations widget, and SDK loader that customers embed. The Worker stripped CSP and served a fake Cloudflare “verify you are human” ClickFix lure (Win+R / Ctrl+V / Enter) downloading malware on Windows; Brevo says that on WordPress sites embedding a Brevo widget, if the visitor was logged in as an administrator, the script also attempted to silently install and activate a plugin. Not affected per Brevo: app.brevo.com, API, email delivery, and customer account data held in Brevo. Remediation: Worker/routes/hostnames removed, key revoked, hardcoded credential removed, Vault + Cloudflare audit alerting planned. Distinct from Brevo’s earlier 9–10 September SSO boundary incident (Trezor phishing wave). Primary: Brevo status write-up. Updated 17 Sep 2026: BleepingComputer reported that Sansec, which first reported the attack, said it may have affected up to 100,000 websites, gave the malicious window as 16:05 to 20:13 UTC, and that BleepingComputer found the plugin, named 'Web Media Optimizer', on VirusTotal. (source: BleepingComputer) Correction (24 Oct 2026): an earlier version paired 'about five and a half hours' with 16:07–20:30 UTC; Brevo's 5h29 is its 15:01–20:30 UTC window. (source: Brevo status write-up)
Brevo status — Cloudflare Worker ClickFix write-up (14 Sep 2026 incident)
breaches cloud
Spain's Agencia Española de Protección de Datos (AEPD) blog (14 September 2026; wired by BleepingComputer and SecurityWeek 16 September) says it received the first notification of a personal-data breach in which the incident was allegedly executed by an AI agent using a known large language model. Per the affected organisation's notification (not yet independently verified by AEPD): the agent searched generic files for flaws, completed a successful login, then autonomously probed the application, modified personal data, and accessed invoices. AEPD stresses the report is from the notifier and must be analysed; use of a named model does not imply the model provider was compromised or that the tool was purpose-built for crime. Relevance for defenders: treat agentic chaining (goal → tools → adapt) as a qualitative speed/scale shift for risk analysis, credential/API hygiene, and detection/containment timing — not only for Spanish controllers. Primary: AEPD blog; wires: BleepingComputer, SecurityWeek.
AEPD — first notified breach allegedly via AI agent (14 Sep 2026)
ai identity
