Multiple security vulnerabilities have been identified in Samba, an open-source implementation of SMB/CIFS protocols used for file and printer sharing. These include missing access checks on reparse point operations, improper blocking of directory modifications in vfs_worm, insecure use of HTTP without validation in group policy certificate enrollment, and remote code execution vulnerabilities in the printing subsystem and SAMR. Red Hat has issued an important security advisory with updates for Red Hat Enterprise Linux 9.4 and related products to address these issues.
On June 11, 2026, the Icarus threat group compromised Klue's backend systems, a market intelligence platform used by hundreds of enterprises to sync competitive battlecard data with CRM environments. The attackers exploited a dormant credential from an abandoned prototype integration to harvest OAuth tokens for Salesforce and Gong. Through automated API calls using Python scripts, the group exfiltrated CRM data including business contacts, price quotes, and sales communications from multiple customer Salesforce organizations. Klue detected the anomalous activity on June 12 and revoked OAuth credentials on June 13. The attackers subsequently launched an extortion campaign starting June 16, demanding victims contact them via Session Messenger within 48 hours.
OpenAM, an open-source identity and access management platform, has an LDAP injection vulnerability in its REST API. The flaw exists in the `_queryId` HTTP query parameter used in user and group query endpoints, where input is not properly escaped before being incorporated into LDAP filter strings. This allows authenticated attackers to inject LDAP metacharacters, potentially enabling user enumeration and blind LDAP injection attacks. The vulnerability affects OpenAM versions prior to 16.1.1 and has a high severity rating with a CVSS score of 8.7. A fix is available in version 16.1.1.
# Vulnerability
`SearchFirstActiveDirectoryRealm.findUserDn()` substitutes the user-supplied username from the login form into an LDAP search filter template (default `cn={0}`) **without escaping RFC 4515 filter metacharacters** (`*`, `(`, `)`, `\`, NUL). Combined with `SearchControls.setCountLimit(1)` on the same call site, this allows three distinct attack primitives:
1. **Authentication confusion** — typing username `*` causes the realm to construct filter `cn=*`, return the first directory entry (typically a privileged account in AD ordering), and attempt bind against that DN with the attacker's password.
2. **Audit log evasion** — payload `bob)(uid=alice` is recorded verbatim in audit logs while the realm searches with the malformed filter, breaking accountability/compliance (SOX, PCI-DSS, ISO 27001).
3. **Directory enumeration** — wildcards and timing differences allow reconnaissance of OU structure and admin group membership.
A repo-wide search for any LDAP escape helper (`escapeLdap`, `encodeFilter`, `escapeFilter`, `ldapEscape`) returns **zero hits** — the defense is not just missing, it was never added.
> **Applicability note:** This realm is opt-in. The shipped default LDAP example (`dist/src/conf/shiro.example.ldap.ini`) uses Shiro's `DefaultLdapRealm` with `userDnTemplate` and is **NOT** affected. However, the realm exists precisely to support Active Directory environments where users log in via `sAMAccountName` and the realm must search for the DN first — the canonical LINE corporate AD-backed SSO scenario. Internal deployments using AD-backed login almost certainly select this realm.
---
## Evidence
**File:** `server-auth/shiro/src/main/java/com/linecorp/centraldogma/server/auth/shiro/realm/SearchFirstActiveDirectoryRealm.java`
**Lines 148–176** on branch `main` @ commit `d64a5151`:
```java
@Nullable
protected String findUserDn(LdapContextFactory ldapContextFactory, String username)
throws NamingException {
LdapContext ctx = null;
try {
ctx = ldapContextFactory.getSystemLdapContext();
final SearchControls ctrl = new SearchControls();
ctrl.setCountLimit(1); // line 156 — returns FIRST match only
ctrl.setSearchScope(SearchControls.SUBTREE_SCOPE);
ctrl.setTimeLimit(searchTimeoutMillis);
final String filter =
searchFilter != null ? USERNAME_PLACEHOLDER.matcher(searchFilter)
.replaceAll(username) // line 162 — RAW SUBSTITUTION
: username; // line 163
final NamingEnumeration result = ctx.search(searchBase, filter, ctrl);
...
```
`USERNAME_PLACEHOLDER = Pattern.compile("\\{0}")`. Default `searchFilter = "cn={0}"`.
### Data flow from HTTP login to vulnerable substitution
| Step | Component |
|------|-----------|
| HTTP login form | `POST /api/v1/login` form field `username` |
| `ShiroLoginService.usernamePassword()` (lines 198–223) | applies `loginNameNormalizer` (Unicode lowercase only — **NOT** LDAP escape) |
| `Subject.login(new UsernamePasswordToken(username, password))` | Shiro hand-off |
| `ActiveDirectoryRealm.doGetAuthenticationInfo` (Shiro core) | calls `queryForAuthenticationInfo0` |
| `SearchFirstActiveDirectoryRealm.findUserDn(factory, upToken.getUsername())` | username flows in **verbatim** |
### Repository-wide escape helper grep
| Search term | Hits |
|-------------|------|
| `escapeLdap` | 0 |
| `encodeFilter` | 0 |
| `escapeFilter` | 0 |
| `ldapEscape` | 0 |
---
## PoC
Self-contained JUnit 5 test using UnboundID `InMemoryDirectoryServer` (in-process, no external LDAP required). Drop into `server-auth/shiro/src/test/java/com/linecorp/centraldogma/server/auth/shiro/realm/LdapInjectionPoCTest.java` and add `com.unboundid:unboundid-ldapsdk:7.0.0` as a test dependency.
> The PoC works by subclassing the realm and overriding `findUserDn()` to capture the actual LDAP filter string sent to the directory — the captured filter is the structural evidence, independent of LDAP server strictness about bind outcomes.
```java
/*
* Copyright 2026 LINE Corporation
*
* SECURITY PoC — NOT FOR MERGE INTO THE MAIN TEST SUITE.
*
* This JUnit class demonstrates the LDAP filter injection in
* SearchFirstActiveDirectoryRealm. Drop into
* server-auth/shiro/src/test/java/com/linecorp/centraldogma/server/auth/shiro/realm/
* Adds the UnboundID LDAP SDK as a test dep.
*/
package com.linecorp.centraldogma.server.auth.shiro.realm;
import static org.assertj.core.api.Assertions.assertThat;
import static org.assertj.core.api.Assertions.assertThatThrownBy;
import javax.naming.directory.SearchControls;
import javax.naming.ldap.LdapContext;
import org.apache.shiro.realm.ldap.JndiLdapContextFactory;
import org.apache.shiro.realm.ldap.LdapContextFactory;
import org.junit.jupiter.api.AfterAll;
import org.junit.jupiter.api.BeforeAll;
Incorrect default permissions in ArubaSign, affecting versions prior to v4.6.6. The vulnerability is caused by the assignment of inappropriate permissions during the software’s default installation, whereby the main executable and other programme files located in C:\Program Files have excessive permissions for the ‘Everyone’ group. This could allow an unprivileged user to replace the main executable and/or its components with a malicious file, thereby enabling the execution of arbitrary code. In the worst-case scenario, if the malicious code is executed with elevated privileges (such as those of Administrator or SYSTEM), the attacker could escalate privileges and gain full control of the system, compromising both security and data integrity.
Craft CMS version 5.0.0-RC1 contains a stored cross-site scripting (XSS) vulnerability in the User Permissions page. The vulnerability arises because user group names are rendered without proper HTML escaping. An attacker with administrative privileges can inject arbitrary JavaScript code via the user group name field, which executes when other users view or edit permissions. This advisory is a duplicate and has been withdrawn in favor of GHSA-g3hp-vvqf-8vw6.
Craft CMS 4.x (>= 4.0.0-RC1, < 4.17.0-beta.1) and 5.x (>= 5.0.0-RC1, < 5.9.0-beta.1) contain multiple stored cross-site scripting vulnerabilities where settings names and field option labels are rendered without sanitization (e.g., via the checkbox.twig template, which used {{ label|raw }}). An authenticated administrator (with allowAdminChanges enabled) can inject malicious payloads into section names, volume names, user group names, global set names, generated field names, checkbox/radio option labels, and custom source labels, causing arbitrary JavaScript to execute in other users' control-panel sessions. Fixed in 4.17.0-beta.1 and 5.9.0-beta.1.
Craft CMS from version 5.0.0-RC1 contains a stored cross-site scripting vulnerability in the User Permissions page where user group names are rendered without proper HTML escaping. Attackers with admin access can inject arbitrary JavaScript via the user group name field that executes when other users view or edit permissions.
CVE-2026-52910 is a use-after-free vulnerability in the Linux kernel's Berkeley Packet Filter (BPF) reuseport mechanism. When a classic BPF (cBPF) program is detached from a reuseport group, it is freed immediately without waiting for ongoing readers to complete, causing potential memory corruption. This flaw can lead to out-of-bounds reads and other memory safety issues during concurrent UDP packet processing. The issue has been fixed by deferring the freeing of the cBPF program until after a Read-Copy-Update (RCU) grace period.
Global law enforcement, including agencies from the Netherlands, Canada, United States, and Germany, coordinated Operation Endgame to disrupt TA569, a prominent cybercriminal group tracked since 2018. The operation targeted SocGholish infrastructure, taking down over 100 servers and domains while remediating 14,971 compromised websites. TA569 pioneered web inject techniques using fake browser updates to distribute malware, often leading to ransomware attacks. The group compromised high-traffic websites across multiple industries, affecting millions of visitors globally. Their attack chains involved traffic distribution systems like Keitaro TDS and ParrotTDS, delivering GhoLoader payloads that could lead to ransomware deployment in enterprise environments. Law enforcement actions included server disruption and website disinfection, significantly impacting the threat actor's operations, infrastructure, and reputation within the cybercriminal ecosystem.
A criminal extortion group known as the Silent Ransom Group bypasses traditional hacking techniques by using social engineering and physical intrusion. They initiate contact with a benign-looking invoice email, followed by phone calls impersonating IT staff to gain remote access via legitimate tools. They steal sensitive data from cloud services and corporate email, targeting especially law firms. When remote tactics fail, they escalate by physically visiting offices posing as IT technicians, using USB sticks to compromise systems. The attack relies on confidence tricks rather than malware or exploits.
A group of cybersecurity executives has urged the Trump administration to lift export restrictions on Anthropic's latest AI models, citing concerns that limiting access could disadvantage U.S. defenders more than adversaries. Anthropic's advanced models, including Fable 5 and Mythos 5, have been taken offline to comply with government directives due to their capability to find and exploit software vulnerabilities. The executives argue that other AI models with similar capabilities exist globally, and restricting access to the best defensive tools could weaken U.S. cybersecurity posture. The U.S. government has imposed export controls as part of broader efforts to vet AI systems for national security risks. The situation reflects tensions between government policy and AI developers over balancing innovation, security, and risk management.
An Iranian threat actor group known as Ababil of Minab left a staging server publicly accessible, exposing over 5 GB of data related to multiple victims including LA Metro. The exposed data includes SQL backups, SCADA configurations, attacker tooling, and victim names. LA Metro confirmed their breach in April, but additional victims were only revealed due to this server exposure. This incident reveals operational security failures by the attacker, leading to unintended disclosure of sensitive information.
A vulnerability in the @angular/service-worker package causes the service worker to strip explicit client-defined request credentials and cache policies during request reconstruction. This results in the browser sending credentials such as cookies or authorization headers on requests where they were explicitly omitted, potentially exposing session data. Additionally, private or non-cacheable resources may be cached, allowing sensitive data to persist in the cache after logout. Applications using affected versions with active Angular Service Worker registrations and asset group patterns matching dynamic routes are vulnerable. Patches are available in versions 22.0.0-rc.2, 21.2.15, 20.3.22, and 19.2.23.
Red Hat has issued a security advisory for MySQL 8.0 addressing multiple unspecified vulnerabilities affecting components such as InnoDB, Information Schema, Group Replication Plugin, Optimizer, DML, and JSON. These vulnerabilities are collectively rated as moderate in severity. The advisory references numerous CVEs with unspecified details and impacts. The update is available for Red Hat Enterprise Linux 8 and its extended life cycle variants.
Parse Server is an open source backend that can be deployed to any infrastructure that can run Node.js. Prior to versions 8.6.80 and 9.9.1-alpha.6, a relation query using the $relatedTo operator could read the membership of a Relation field even when that field was hidden from the requesting client by protectedFields, and even when the object owning the relation was not readable by the client under its ACL or class-level permissions. The request requires only the public API credentials that Parse clients normally carry — no user session, master key, or Cloud Code is needed. As a result, an unauthenticated client who knows or obtains the owning object's objectId could enumerate the objects linked through a protected relation, or combine the operator with an objectId constraint to use it as a membership oracle — confirming whether a specific object is linked to a private parent. This affects applications that rely on protectedFields or object ACLs to keep Relation membership confidential, such as private group memberships, block lists, or account-to-resource associations. This issue has been patched in versions 8.6.80 and 9.9.1-alpha.6.
The UNC1151/Ghostwriter group is conducting high-intensity phishing campaigns targeting Gmail accounts of Polish citizens since March 2026. The campaigns primarily target individuals in political and public life, prominent positions, researchers, journalists, public administration and law enforcement employees, and their associates. Attackers use fraudulent emails impersonating Gmail administrators, claiming suspicious activity or policy violations to pressure victims into verifying their accounts. The phishing infrastructure captures login credentials and two-factor authentication codes through fake login panels. The group utilizes dedicated domains, Netlify subdomains, and compromised websites to host phishing pages. Campaigns run primarily on weekdays with new domains appearing almost daily, demonstrating persistent operational tempo against Polish targets.
A privilege escalation vulnerability in Palo Alto Networks PAN-OS® software allows an authenticated administrator with access to the Command Line Interface (CLI) to perform actions on the device with root privileges.
The security risk posed by this issue is significantly minimized when CLI access is restricted to a limited group of administrators and by restricting access to the management interface to only trusted internal IP addresses according to our recommended best practice deployment guidelines https://live.paloaltonetworks.com/t5/community-blogs/tips-amp-tricks-how-to-secure-the-management-access-of-your-palo/ba-p/464431 .
This issue is applicable to PAN-OS software on PA-Series and VM-Series firewalls and on Panorama (virtual and M-Series).
Cloud NGFW, and Prisma® Access are not impacted by this vulnerability.