>
av.Amit VijayanCYBERSECURITY & RESEARCHLet\u2019s connect
Back to the library

Cyber news

Daily Cyber Threat Brief — November 16, 2026: La Poste subsidiary breach exposes 22 million people in France

Thumbnail for: Daily Cyber Threat Brief — November 16, 2026: La Poste subsidiary breach exposes 22 million people in France

Case file — Lead incident: La Poste has confirmed that its business-mail subsidiary Maileva lost the names and postal addresses of 22 million people — roughly a third of France's population — in a mid-September attack on its production mail platform. The victims are not Maileva account holders: they are the recipients of letters, payslips, invoices, and registered mail sent through the platform by Maileva's business customers, including social-housing providers and medical laboratories. The attackers reached a mail-tracking database holding about 39 million mail references through an application-layer vulnerability; deduplicated, those references cover 22 million distinct people. Maileva says no bank details were compromised and the contents of letters were not confirmed stolen, though notices from individual customers show some letters carried invoice numbers, care dates, and amounts due. The breach was first flagged by the specialist sites FrenchBreaches and Cyberattaque.org in October; La Poste confirmed the scale to AFP on November 14. Maileva has notified France's data protection authority, the CNIL, filed a criminal complaint, and a judicial investigation is underway. Status: services were suspended on September 23 and restored by September 29; the 22 million people affected face phishing and fake payment-collection scams built on their stolen names and addresses.

Incident timeline

DateDevelopment
September 9, 2026First suspicious activity in Maileva's production environment, according to Docaposte's internal fact sheet; the intrusion is detected on September 16–17.
September 23, 2026, 7:00 a.m.Maileva suspends its services after detecting unauthorized attempts to access customer accounts; full operations resume September 28–29.
October 5, 2026An internal Docaposte client fact sheet describes the attack as an application-layer vulnerability in the production information system and puts the affected population at 22 million people.
October 9, 2026The breach-monitoring site FrenchBreaches surfaces the Docaposte fact sheet; Cyberattaque.org corroborates it.
November 8, 2026A forum actor using the alias "SmithBianl" posts a 499-row sample of data allegedly stolen from TSMC and opens an auction for a claimed 6.9GB dataset of 46,000 employee records, starting at 3 BTC.
November 12, 2026, ~04:00 KSTAn attacker breaks into servers run by MetaM, the contractor that handles customer support for South Korean crypto exchange Coinone; MetaM alerts Coinone at about 09:20 the same morning. TSMC files a Form 6-K with the SEC confirming employee data was posted online without authorization.
November 13, 2026The FBI releases FLASH-20261113-001, a TLP:CLEAR advisory warning that the ARTEX agentic AI framework is being weaponized to automate intrusions. Maileva confirms the 22 million figure to Cyberattaque.org.
November 14, 2026La Poste confirms the Maileva breach to AFP: 22 million people affected, the largest breach acknowledged in France. MetaM reports its breach to the Korea Internet & Security Agency.
November 15, 2026BleepingComputer reports the TSMC breach; Coinone publicly discloses the MetaM intrusion; the FBI's ARTEX warning reaches the wider press. The actor's one-week deadline for releasing the TSMC data for free falls due.

1. Maileva breach exposes names and addresses of 22 million people in France

La Poste has confirmed one of the largest data breaches ever acknowledged in France: a mid-September attack on its subsidiary Maileva exposed the first names, last names, and postal addresses of 22 million people. Maileva, part of La Poste's Docaposte digital division, prints and posts documents — payslips, invoices, registered letters, meeting notices — on behalf of more than 70,000 business customers, processing 180 million documents in 2024. The 22 million affected are the recipients of that mail, not Maileva's own customers, and many will never have heard of the company: they include patients of medical laboratories in the Inovie group, tenants of social-housing providers such as Flandre Opale Habitat and Altéal, and recipients of hospital correspondence such as Hôpital Foch in Suresnes.

According to an internal Docaposte fact sheet dated October 5, first surfaced by FrenchBreaches on October 9, the attackers exploited a vulnerability in the application layer of Maileva's production information system and reached a mail-tracking database holding about 39 million mail references. Because one person can receive several letters from different senders, those references collapse to about 22 million distinct people once duplicates are removed. The first suspicious activity dates to September 9; Maileva detected the intrusion on September 16–17, suspended services at 7 a.m. on September 23 after unauthorized attempts to access customer accounts, and restored full operations on September 28–29.

Maileva says bank details were not compromised, and the theft of the full contents of letters has not been confirmed. That reassurance has limits: notices from individual customers show that some letters the attackers could reach contained genuine invoice numbers, dates of care, and amounts due. The company has notified the CNIL, filed a criminal complaint, and says a judicial investigation is underway; La Poste says all affected client companies have been informed, though that does not establish that all 22 million people have been reached individually. The pro-Russian hacking group Noname057 has claimed responsibility for the attack, according to Security Affairs; La Poste and Maileva have said nothing about the perpetrators, so the claim remains unconfirmed.

Investigation notes

Why this is the lead: 22 million people is about a third of France's population, and the victims are ordinary letter recipients — patients, tenants, hospital patients — who had no relationship with the breached company and no way to opt out. The data is exactly what fuels convincing phishing and fake debt-collection scams: real names tied to real addresses and, in some cases, real invoice and care details.

What is confirmed: the 22 million figure and the data types (La Poste to AFP, November 14; Maileva to Cyberattaque.org, November 13); the application-layer cause and the September 9 first access, September 16–17 detection, and September 23 suspension (Docaposte fact sheet, October 5); CNIL notification, criminal complaint, and judicial investigation.

What is not known: the identity of the attackers — the Noname057 claim is reported, not confirmed — whether any of the data has been sold or published, and whether letter contents beyond the tracking data were taken. Nothing establishes at this stage that the files have been leaked or put up for sale.

What recipients should do: treat any unexpected letter, email, or call citing a real invoice number, care date, or amount due with suspicion, and verify payment demands through the sender's own channel rather than a link or number in the message. Maileva customers should check the notices their own providers have issued, since the exposure varies by client.

2. TSMC employee data breach linked to third-party cloud provider

Taiwan Semiconductor Manufacturing Company (TSMC) has confirmed that personal employee data was leaked online after a threat actor put 6.9GB of information allegedly stolen from the company up for auction. On November 8, someone using the alias "SmithBianl" posted a 499-row text file on a hacker forum as a sample of a dataset the actor claims holds about 46,000 employee records, including national identification numbers, residential addresses, contact details, and banking information. The actor offered the full dataset for auction with a starting bid of 3 BTC — approximately $65,000 — and warned that it would be leaked for free within a week if no buyer appeared, setting the first deadline to November 15.

TSMC verified the leaked samples and confirmed they matched data stored in the systems of a cloud service provider it uses, pointing to a breach outside its own infrastructure. In a Form 6-K filed with the US Securities and Exchange Commission on November 12, signed by CFO Wendell Huang, the company said it had found no evidence that its own systems were compromised, and that the incident does not affect its manufacturing operations or its customers, including Apple and Nvidia. The company is notifying impacted current and former employees, providing support, and has engaged a third-party cybersecurity firm to help manage the situation. The cloud provider involved has not been named.

Investigation notes

The pattern to note: this is the third story in this brief where the breached system belonged to a supplier, not the named victim — TSMC's cloud provider, Coinone's support contractor, and Maileva's mail platform all sit between the organization and the people whose data was taken. Third-party exposure is the week's theme, not a coincidence.

What is confirmed: the 499-row sample, the 6.9GB / 46,000-record claim, the 3 BTC starting bid, and the November 15 deadline (forum post via KELA, reported by BleepingComputer, November 15); TSMC's verification that the samples match data held by a cloud service provider and its SEC filing of November 12.

What is not known: which cloud provider was breached, how the data was taken from it, whether the auction found a buyer, and whether the full dataset has been released now that the deadline has passed. The 46,000-record figure is the actor's claim; only the sample has been verified.

What TSMC staff should do: current and former employees should watch for the company's notification and treat any contact citing their national ID number or banking details as a phishing attempt. The incident is a reminder to check what employee data sits in third-party systems, and what those providers are contractually required to report.

3. Coinone breach exposes passwords and ID numbers through hacked support vendor

South Korean cryptocurrency exchange Coinone has disclosed that an attacker broke into the servers of MetaM, the contractor that runs its customer consultation service, and viewed or removed personal data customers supplied during chat and phone support sessions. The intrusion happened at about 04:00 KST on Thursday, November 12; MetaM told Coinone at about 09:20 the same morning, and Coinone published its public notice at 08:00 KST on November 15 — roughly 76 hours after the attack began.

The notice lists eleven categories of information that may have been exposed: phone numbers, names, email addresses, dates of birth, bank account numbers, passwords, resident registration numbers, identification document issuance dates, residential addresses, cryptocurrency wallet addresses, and transaction identifiers. Coinone is explicit that this is the union of everything that could have been exposed, not a per-customer list — what leaked for any one person depends on what they shared during their own support conversation, and customers who never contacted support are outside the scope. The exchange has not said how many customers are affected. Its trading, account, and wallet systems run separately from MetaM's servers, and it says it has found no sign they were touched; logins require a second factor through an OTP or KakaoTalk verification on the customer's registered phone. Coinone has blocked the attacker's access routes, switched its outsourced support to email only, and reported the incident to the Korea Internet & Security Agency, the Personal Information Protection Commission, and the Financial Supervisory Service.

Investigation notes

Why the field list matters: four of the eleven categories are identity anchors that cannot be reissued at will — name, date of birth, resident registration number, and ID document issue date — and four more carry immediate fraud value: password, bank account number, wallet address, and transaction ID. Support chats are where customers overshare, and this breach shows what that costs when the desk, not the exchange, is the target.

What is confirmed: the intrusion time and notification chain (Coinone notice, November 15; MetaM's own notice gives a slightly different starting point, an outage spotted at 06:03 on November 12, and a KISA report at 03:44 on November 13); the eleven field categories; the regulator notifications; the email-only support switch.

What is not known: how many customers are affected, how the attacker got in, whether the records were encrypted at rest, and whether the data has been used. Coinone says the leaked information alone is not enough to log in or withdraw assets — but it is exactly enough to run convincing impersonation scams, and the exchange warns customers it never asks for passwords, OTPs, or verification codes by phone, text, or email.

What Coinone customers should do: change the password used with Coinone, watch for phishing that quotes details from a real past support conversation, and treat any request to withdraw coins or change a wallet address as a scam.

4. FBI warns ARTEX AI framework is automating intrusions in under three hours

The FBI has warned that hackers are using ARTEX, an open-source AI-powered penetration testing framework, to automate cyberattacks against organizations worldwide — including a recent ConnectWise supply-chain breach that compromised Microsoft 365 customer environments. In a FLASH advisory released on Friday, November 13 (FLASH-20261113-001, TLP:CLEAR), the bureau said multiple threat actors have weaponized the framework since at least August 2026 to target organizations across financial services, government, technology, defense, semiconductors, energy, healthcare, manufacturing, telecommunications, and legal sectors.

ARTEX is an autonomous, multi-agent penetration testing platform written in Go that translates high-level goals described in natural language into machine commands executed by specialized agents, with minimal human interaction beyond setting the objective. It integrates with external large language models — Anthropic Claude, OpenAI GPT, DeepSeek, GLM, Grok, and Kimi — and the FBI says it lets actors with minimal technical expertise run complex, multi-target operations simultaneously. The speed is the headline risk: citing Microsoft, the FBI notes some intrusions progressed from initial targeting to privileged access in under three hours, leaving defenders little time to coordinate a response even when alerts fire. Notable campaigns include financially motivated attacks against South Korean financial institutions that exfiltrated personal information from approximately 67,000 individuals, and the ConnectWise Cloud Backup supply-chain compromise that gave attackers access to more than 2,250 downstream tenants.

ARTEX was originally released on GitHub in July 2026 by a developer known as Autumn-27 and reached an estimated 1,500–2,000 stars within ten weeks. After reports emerged that threat actors were abusing it, the developer removed the repository on October 8 — but forks, archived copies, localized variants, and a commercial ARTEX PRO version keep it available. As of early November, researchers had identified more than 1,600 publicly exposed ARTEX consoles, with the actual number of deployments believed to be significantly higher. The FBI's 13-page advisory includes indicators of compromise and a prioritized list of mitigations, topped by inventorying and shrinking the external attack surface; cybersecurity researcher MalwareMustDie released an open-source ARTEX detection signature pack on Friday to help defenders identify the attacks.

Investigation notes

Official anchor: the FBI FLASH advisory of November 13, built on eleven industry and government sources including Microsoft, Google Threat Intelligence Group and Mandiant, Akamai, Censys, CrowdStrike, Zscaler ThreatLabz, CERT-EU, and Check Point Research. The advisory is TLP:CLEAR and may be shared without restriction.

What is confirmed: weaponization by multiple actors since at least August 2026; the under-three-hours intrusion timeline (Microsoft); the South Korean bank campaign (~67,000 people) and the ConnectWise compromise (2,250+ downstream tenants); more than 1,600 exposed ARTEX consoles as of early November (Censys).

What is not known: how many organizations have been hit in total, and how many of the exposed consoles are attacker-controlled versus legitimate test deployments. The FBI cautions that IP addresses in its IOC list may already have changed. The framework's speed, not its novelty, is what the advisory treats as the change: ARTEX generates high volumes of noisy traffic that may trigger alerts, but the attacks complete before a coordinated response can form.

What defenders should do: start with the FBI's top recommendation — inventory internet-exposed applications, self-service portals, partner and supplier systems, and legacy services, and shrink that surface. Then apply the FLASH mitigations and the open-source ARTEX detection signatures, and treat partner and supplier portals as the entry points ARTEX agents have used most.

Sources

Keep exploring.

All articles