The internet is broken right now.
105 critical CVEs without a fix in the last 48 hours. Somewhere a staging server is public. It always is. 33 more already have a fix that someone still has to apply.
-
CVE-2026-126610
Progressive Robot hMailServer 6.3.4 through 6.3.7, with a MySQL or MariaDB database and TlsMode=verify-full, does not hand the database client the CA file that TlsCaFile names when the file's name contains a character outside ASCII on Linux, or above U+00FF on Windows, which allows an attacker who can intercept the connection to the database server to read and modify the traffic between hMailServer and its database, by presenting, with MariaDB Connector/C 3.4.2 to 3.4.10 as the client, a certificate from any authority, its own included, for any name, and otherwise a certificate for the database server's name issued by any certificate authority in the system's trust store. The server converted the name for the client with the C runtime's conversion in the C locale, which gives an empty string for such a name, and then set no CA file at all, without an error, while certificate verification stayed on, so the client - MariaDB Connector/C, bundled on Windows and the distribution's on Linux - verified the certificate of a database server over the network against the system's trust store instead of the configured CA; Connector/C 3.4.2 to 3.4.10 (bundled on Windows in 6.3.4 to 6.3.6) went further and, with no CA file, took a certificate within its dates whose issuer it could not find, for any name, with neither its chain nor its name checked and no password proof, when the server offered first a sign-in method that hashes the password, which an attacker in the middle chooses (MariaDB's CONC-853, fixed in Connector/C 3.4.11), and Connector/C 3.4 accepted from a MariaDB 11.4 or later server a certificate the store did not vouch for when that server proved it held the account's password; on a connection it treats as local (127.0.0.1, ::1, a Unix socket, a named pipe, and localhost on Windows), Connector/C 3.4.1 and later checked neither the certificate's issuer nor its name, so that a local attacker able to stand in for the database server could present any certificate within its dates of its own making. The settings line of hmailserver --check-database still named the configured file, and the application log said the certificate had been checked. With MariaDB Connector/C 3.4.1 to 3.4.9 as the client - bundled on Windows in 6.3.4 to 6.3.6, and the client of Linux distributions that have not patched it - the missing CA file also met the precondition of CVE-2026-107819, so that a certificate issued for any other name by a trusted authority was enough to obtain the database password. On Linux the file was found only when the server ran on glibc under a locale whose character set holds the name's characters (a UTF-8 locale, or a legacy 8-bit one the system has, in whose bytes the file is then named); with no locale, LC_ALL=C or a locale the system lacks, and on Alpine (musl), the connection failed instead. The configuration is not the default: TlsMode defaults to prefer, which checks no certificate.
-
CVE-2026-121392
FreeRDP before 3.32.2 contains an improper certificate validation vulnerability in tls_verify_certificate() that accepts Common Name fallback despite non-matching SAN entries and matches IP literals against DNS strings. Man-in-the-middle attackers holding a certificate from a client-trusted CA can bypass server identity checks to intercept RDP sessions, including keystrokes, clipboard, and screen content.
-
CVE-2026-106036
Szafir accepts any client or server certificate without performing certificate and hostname validation, allowing HTTPS connections without authenticating the remote server. An attacker can provide an arbitrary SSL/TLS certificate and cause the application to download an update file from their server. This issue was fixed in version 2.0 build 708.
-
CVE-2026-116474
The `GITEA_TOKEN` of a Gitea Actions job authenticates as the `gitea-actions` system user, and the `/api/v1/user` email and GPG key endpoints accepted that user, so a job could register an email address and a GPG key on it; without required email confirmation (the default) the address is stored as activated. Commit verification treats a key without a real owner as the instance signing key, so any user who can run a workflow job could make commits signed with their own key show as trusted and signed by the instance in every repository where they appear, including pull requests to repositories the user cannot write to. The impact is limited to the displayed signature status: enforcement such as required signed commits only checks that a signature is verified and is unaffected, and no access is gained. The `/api/v1/user` routes now reject system users.
-
CVE-2026-107567
Use of a less trusted source in the DKIM signing of Progressive Robot hMailServer 6.0.0 through 6.3.6 allows a remote unauthenticated attacker to have mail they write signed with the DKIM key of a domain the server hosts, so that it passes DKIM and DMARC for that domain at other receivers. With DKIM signing enabled for a hosted domain, the server signed every message on its first delivery pass as the domain named in its From header, before local delivery, whatever the message's origin and without checking that its sender had authenticated. A sender anywhere could therefore send the server, as ordinary inbound mail, a message whose From is any address at such a domain, and the server signed it with that domain's key. Every copy the server then sent on to an address outside carried a valid signature aligned with the forged From, whatever the domain's DMARC policy: an account's forward, a mail rule's forward or copy, a Sieve redirect, an alias or distribution list with an outside member, the mirror copy, a managed-list or journal copy, or a message released from quarantine. A signed copy can also be re-sent to other recipients while the domain's key is unchanged. Exploitation needs DKIM signing to be enabled for the hosted domain (it is off until configured) and, for the forgery to leave the server, a forward, alias, list or rule that carries the domain's mail to an outside address. The signer likewise signed a message submitted over authenticated SMTP as whatever hosted domain its From header named, a header authenticated SMTP does not hold to the submitting account, so that an account of one hosted domain could have mail written in the name of another hosted domain signed with that domain's key; this mattered where the server holds signed-in senders to their own envelope addresses (SmtpAuthenticatedSenderCheck, off by default), as the signature was then the remaining way for such mail to pass the other domain's DMARC policy.
-
CVE-2026-116081
Sleipnir Mobile for iPhone / iPad and Sleipnir Mobile for iPhone / iPad Black Edition provided by Fenrir Inc. are vulnerable to improper certificate verification. A man-in-the-middle attack may allow an attacker to eavesdrop on an encrypted communication.
-
CVE-2026-121565
Tanium addressed an improper certificate validation vulnerability in multiple products. Products have been updated to distrust several intermediate CA certificates associated with compromised private keys.
-
CVE-2026-48057
Botan is a C++ cryptography library. Prior to 3.13.0, Botan's X.509 Distinguished Name decoder flattens the Attribute-Value Assertions in each Relative Distinguished Name and loses multi-value RDN grouping before DN name-constraint enforcement. An attacker controlling a name-constrained intermediate CA can issue a certificate with a multi-value RDN whose first Common Name matches the permitted constraint and whose second Common Name is attacker-chosen, causing the constraint check to accept the flattened first value. If the certificate contains no subjectAlternativeName extension, the second Common Name can then be accepted as a hostname; Botan does not fall back to Common Name hostname matching when any subjectAlternativeName is present. The exposure is limited to deployments using DN name constraints, which are uncommon outside internal PKIs, and to certificates without a subjectAlternativeName. This issue is fixed in version 3.13.0.
-
CVE-2026-111494
Incorrect validation of TLS certificates in the MatchPoint Android app due to a TrustManager that accepts any certificate and a permissive hostname validator. Additionally, the app allows unencrypted HTTP traffic. An attacker capable of intercepting the victim's traffic could read and modify communications, including credentials and personal data.
-
CVE-2026-103330
The Sadad QA Payment WordPress plugin through 1.28 does not correctly verify the authenticity of incoming payment notifications, as the verification is keyed on a merchant secret that is empty by default and the rest of the signed data is attacker-controlled, allowing unauthenticated attackers to forge a valid confirmation and mark orders as paid without payment.
-
CVE-2026-123590
Hickory DNS hickory-net 0.26.0-beta.1 before 0.26.2 and hickory-proto through 0.26.0-alpha.1 accept NSEC and NSEC3 records for DNSSEC nonexistence proofs when any other record with the same owner name passes signature validation. Attackers able to supply DNS responses can mix validly signed records with bogus same-name NSEC or NSEC3 records of different types in the authority section to forge nonexistence proofs.
-
CVE-2026-84918
Improper Verification of Cryptographic Signature vulnerability in Drupal Monobank payment API allows Signature Spoof. This issue affects Monobank payment API versions: from 0.0.0 to 1.0.3.
-
CVE-2026-123613
Rustls from 0.23.13 before 0.23.45 accepts TLS 1.3 handshake messages at the wrong encryption level when they follow a key-changing message in the same record, because Deframer::aligned does not enforce RFC 8446 record-boundary alignment. A peer or on-path attacker can pack a plaintext EncryptedExtensions with the ServerHello, which rustls accepts unencrypted.
-
CVE-2026-61813
ModSecurity is an open source, cross platform web application firewall (WAF) engine for Apache, IIS and Nginx. Prior to 2.9.15 and 3.0.17, the HttpsClient::download function in src/utils/https_client.cc sets libcurl CURLOPT_SSL_VERIFYHOST to 1 instead of the strict value 2L. Under adversary-in-the-middle and TLS interception conditions, this non-strict setting may weaken TLS peer identity validation and increase the risk of accepting a certificate for an unintended host. End-to-end exploitation is non-trivial in typical environments. This issue is fixed in versions 2.9.15 and 3.0.17.
-
CVE-2026-112265
LangBot through 4.11.0 contains a missing authenticity verification vulnerability that allows unauthenticated attackers to inject forged Lark/Feishu events by skipping signature checks. Attackers knowing a bot UUID can POST fabricated im.message.receive_v1 or card.action.trigger events to drive the LLM pipeline, tools, and approvals as other users.
-
CVE-2026-105996
The Booking Package WordPress plugin before 1.7.31 does not generate booking cancellation tokens using a secure source of randomness, deriving them instead from values it exposes publicly, so an unauthenticated attacker can reconstruct an arbitrary booking's token, disclose that booking's personal data and, where visitor cancellation is enabled, cancel the booking.
-
CVE-2026-6930
IBM Concert 1.0 through 3.0.0 uses weaker than expected cryptographic algorithms that could allow an attacker to decrypt highly sensitive information.
-
CVE-2026-115906
zip is a zip library for Rust that supports reading and writing simple ZIP files. From 0.6.0 until 9.0.0, the AES entry reader in src/aes.rs verifies its HMAC only when the attacker-controlled data counter reaches zero. If the underlying bounded reader reaches the real end of the archive while data_remaining is still positive, or if an encrypted entry has no payload, the reader can return a clean end of stream without reading and comparing the authentication tag. Applications using by_index_decrypt, by_name_decrypt, by_path_decrypt, or password-enabled by_index_with_options can therefore accept modified plaintext without an authentication error. This issue is fixed in version 9.0.0.
-
CVE-2026-108013
BigFix Platform uses Mozilla CA certificate bundle (PEM). Versions of the bundle dated before 2026/10/20 still trust certificates whose private keys have been compromised. When BigFix Platform validates servers using such a bundle, an on-path attacker can impersonate remote servers and tamper with files downloaded by the application. The issue is resolved by using bundle version 2026/10/20 or later. The fixed bundle is available starting from BES Support v1520.
-
CVE-2025-13833
A vulnerability exists in SecureConnector for Windows when the “Additional Appliance Connections” setting is enabled and the agent’s configuration file contains more than three trusted‑server entries. Some entries omit TLS certificate fingerprints, which allows the agent to accept and establish sessions with a malicious server. An attacker who controls the wired or wireless network can redirect traffic to a malicious server. The attacker can then use the agent’s built‑in command/file features to execute arbitrary commands as SYSTEM, granting full administrative control and the ability to disable endpoint protections or exfiltrate data.
-
CVE-2026-88998
The Tickera WordPress plugin before 3.6.0.4 does not correctly verify the integrity of incoming payment notifications, and the validated signature does not cover the order or the amount, allowing unauthenticated attackers to forge a payment confirmation and mark arbitrary orders as paid without paying.
-
CVE-2026-17309
IBM API Connect 12.1.0.0 through 12.1.1.2 could allow a remote attacker to obtain sensitive information and tamper with configuration due to improper certificate validation.
-
CVE-2026-108910
IBM Db2 Genius Hub 1.1.2, 1.1.3, 1.1.4, 1.1.5 could allow a remote authenticated attacker to obtain sensitive information due to missing required cryptographic step.
-
CVE-2026-123271
ChatCLI before v1.215.2 contains an improper validation vulnerability in the MCP OAuth login flow that uses server-supplied Protected Resource Metadata as the resource indicator without checking it matches the configured server URL. A malicious MCP server can declare another resource on the same authorization server, obtain the user's bearer token for it, and replay that token against the other resource.
-
CVE-2026-84790
OpenVPN 2.7_alpha1 through 2.7.7 allows remote authenticated users to impersonate others via a crafted certificate subject field
-
CVE-2026-17308
IBM API Connect 12.1.0.0 through 12.1.1.2 could allow a remote attacker to obtain sensitive credentials and manipulate tool responses due to improper certificate validation.
-
CVE-2026-115758
SumatraPDF is a multi-format reader for Windows. In version 3.6.1 and earlier, HttpGet() in src/base/Http_win.cpp unconditionally passes INTERNET_FLAG_IGNORE_CERT_CN_INVALID to WinINet, disabling hostname matching for HTTPS update metadata and crash-symbol requests while retaining certificate-chain validation. A network-positioned attacker with a Windows-trusted certificate for an attacker-controlled domain can impersonate the project's update host, inject promo text and a clickable URL before update metadata validation, and provide an attacker-controlled archive to ExtractSymbols(). Installer replacement is not directly possible because HttpGetToFile() validates TLS and MaybeStartUpdateDownload() restricts the download host. No fixed version is available as of this review.
-
CVE-2026-118765
Botan is a C++ cryptography library. From 3.7.0 until 3.13.0, an application that calls AutoSeeded_RNG::clear() and then calls randomize with an empty output buffer while the generator is unseeded can cause AutoSeeded_RNG to mark itself seeded without consulting the configured entropy source. On builds without BOTAN_HAS_SYSTEM_RNG, the fallback uses only a clock value and process ID, so a later non-empty request can produce output from predictable state. This issue is fixed in version 3.13.0.
-
CVE-2026-118372
LibreSSL ocspcheck through 4.4.0 contains a CRLF injection vulnerability that allows attackers to inject HTTP headers by embedding CR/LF sequences in a certificate's AIA OCSP URI path. Because url2host() copies the path without rejecting control characters and http_open() places it in the POST request line, attackers can inject headers or split OCSP requests sent to the responder host.
-
CVE-2026-103301
The Payment Gateway for Telcell WordPress plugin before 2.0.5 does not verify the authenticity of incoming payment notifications, because the secret used to sign them is empty by default, allowing unauthenticated attackers to forge a valid notification and mark arbitrary WooCommerce orders as paid. The same handler also marks an attacker-chosen order as failed when given an invalid signature.
-
CVE-2026-116641
The DB Tools rely on the TLS implementation in Golang’s standard library, which generally uses the system CA store. However, the crypto/x509 package also has its own internal denylist of known-bad CA certificates. Golang 1.27.3 adds the eMudhra certificates to this list. DB Tools version 100.19.0 was compiled with Golang 1.26.8, which does not include this check. The issue is resolved by updating to Golang 1.27.3 in DB Tools 100.20.0
-
CVE-2026-115930
The MongoDB Shell (mongosh) relies on Node.js as a third-party runtime that carries its own TLS root certificate store. This certificate store has the recoverable eMudhra intermediate CA certificates. This issue impacts MongoDB Shell (mongosh) versions prior to 2.14.0 and can be resolved by updating to mongosh 2.14.0 which bundles Node.js 24.22.1 or later. If mongosh is installed through npm, it uses the Node.js runtime installed on the host; update that runtime separately.
-
CVE-2026-116189
rclone from 1.55.0 before 1.75.2 contains a signature verification bypass in stable selfupdate that trusts checksums from unsigned bytes preceding the PGP-clearsigned SHA256SUMS block. Attackers controlling the release distribution path can prepend an unsigned checksum line matching a malicious archive, causing selfupdate to install an attacker-controlled rclone executable.
-
CVE-2026-54372
VitaFit Android application (com.vt.vitafit) before 2.8.1.0 contains an insecure TLS/SSL certificate validation vulnerability that allows network-based attackers to intercept and modify HTTPS traffic by exploiting a custom X509TrustManager that trusts all certificates and a hostname verifier that accepts any hostname. Attackers can perform a man-in-the-middle attack to capture account sessions and health-related API data, and use a captured session to delete the account's authentication record and prevent legitimate login.
-
CVE-2026-87947
Improper Verification of Cryptographic Signature vulnerability in Drupal SAML SSO - Service Provider allows Signature Spoof. This issue affects SAML SSO - Service Provider versions: from 0.0.0 to 3.2.0.
-
CVE-2026-122098
CAI Content Credentials is affected by an Improper Certificate Validation vulnerability that could result in a Security feature bypass. An attacker could leverage this vulnerability to bypass security measures and gain unauthorized write access. Exploit depends on conditions beyond the attacker's control. Exploitation of this issue does not require user interaction.
-
CVE-2026-87944
Improper Certificate Validation vulnerability in Drupal SAML SSO - Service Provider allows Adversary in the Middle (AiTM). This issue affects SAML SSO - Service Provider versions: from 0.0.0 to 3.2.0.
-
CVE-2026-61810
libspdm is a sample implementation that follows the DMTF SPDM specifications. In version 1.4.0, both the FINISH request (Table 80) and the FINISH_RSP response (Table 81) gained two new fields, OpaqueDataLength and OpaqueData, placed before the trailing signed/authenticated field. The transcript definitions used to compute and verify the FINISH signature, the FINISH HMAC (RequesterVerifyData), and the FINISH_RSP HMAC (ResponderVerifyData) were not updated and still terminate with a step of the form. Per DSP0274, the "SPDM Header Fields" are only the four header bytes (SPDMVersion, RequestResponseCode, Param1, Param2). A literal reading therefore excludes OpaqueDataLength and OpaqueData, even though those fields are transmitted before the trailing signature/HMAC field. Because the signature/HMAC must cover every field transmitted before it to "protect the entire handshake against manipulation by an active attacker," an active attacker could otherwise modify the opaque data in flight without invalidating the signature or HMAC. This issue has been patched in version 1.4.1.
-
CVE-2026-91908
Dell OpenManage Enterprise, versions prior to 4.8.0, contains an Improper Verification of Cryptographic Signature vulnerability. A high privileged attacker with adjacent network access could potentially exploit this vulnerability, leading to Code execution.
-
CVE-2026-118776
Forge (also called node-forge) is a native implementation of Transport Layer Security in JavaScript. Prior to 1.4.1, the affected implementation used decipher.finish(), CBC and ECB modes, PKCS#7 padding to process the relevant input or security decision. node-forge validated only the final byte of PKCS#7 padding during CBC and ECB decryption and accepted zero, oversized, or mismatched padding bytes. Malformed ciphertext could consequently be accepted and up to four plaintext blocks could be silently removed, causing an integrity failure in applications that treated successful decryption as proof of valid padding. This issue is fixed in version 1.4.1.
-
CVE-2026-107692
The War Thunder game client transmits its authentication bearer token (auth_jwt) in the very first data packet after the TCP handshake, using only LZ4 compression rather than encryption. An attacker positioned on the network path can passively intercept and decompress this packet to reveal the plaintext JSON payload. This vulnerability allows for temporary authenticated impersonation, enabling an attacker to access game services and certain user data (such as email address and displayed phone number) as the affected user. While this allows for unauthorized access to gameplay, it does not facilitate full account takeover, as it cannot be used to modify account security settings, change passwords, bypass two-factor authentication for account management, or access financial information.
-
CVE-2026-123591
Hickory DNS hickory-net 0.26.0-beta.1 before 0.26.2 and hickory-proto through 0.26.0-alpha.1 fail to verify, per RFC 6840 §4.4, that an NSEC or NSEC3 record denying a DS RRset also proves the delegation's NS record exists. Attackers can reuse child-side NSEC or NSEC3 records to fake an insecure delegation and forge child-zone records.
-
CVE-2026-120423
undici is an HTTP client for Node.js. In versions 7.15.0 through 7.30.1 and 8.0.0 through 8.12.3, fetch verifies Subresource Integrity metadata by splitting the integrity string on a single space, but the SRI grammar separates metadata by ASCII whitespace, which also includes the horizontal tab. An integrity value whose digest is preceded by a tab or other ASCII whitespace is parsed as a single unrecognized token and skipped, and when no recognized algorithm remains undici treats the integrity list as empty and accepts the response, so a server returning a body different from the pinned digest passes the check. An application that pins a response with a valid but whitespace-surrounded SRI value receives attacker-controlled bytes with no integrity error, and the attacker needs only to control the response body, not the integrity metadata. The issue is fixed in undici 7.30.2 and 8.12.4, and users should upgrade to one of those versions or later. As a workaround, supply SRI integrity metadata without surrounding whitespace, using a single space to separate multiple digests.
-
CVE-2026-123538
OpenAM before 16.1.4 does not verify that a signed SAML 1.x assertion or response was signed by the partner named as its issuer, accepting a signature from any certificate in OpenAM's keystore. An attacker holding such a private key, such as a trusted SAML 1.x partner, can forge assertions claiming another partner's issuer and sign in as that partner's users, including privileged ones.
-
CVE-2026-17393
IBM API Connect 10.0.8.0 through 10.0.8.12 could allow a local attacker to obtain sensitive information due to the use of a cryptographically weak pseudo-random number generator.
-
CVE-2026-121624
Nanosystems Supremo for Windows and macOS before 5.0.2.3874 contains an improper certificate validation vulnerability that allows adjacent attackers to intercept TLS connections by presenting untrusted certificates. Attackers on the same network can terminate and relay the client's TLS session using self-signed certificates to read usernames, SHA-256 password hashes, and license codes.
-
CVE-2026-14434
IBM Turbonomic kubeturbo agent 8.16.0 through 8.21.2 does not validate or improperly validates TLS certificate validation, which could allow an attacker to obtain sensitive information using man in the middle techniques.
-
CVE-2026-14435
IBM Turbonomic kubeturbo agent 8.16.0 through 8.20.3 does not validate or improperly validates TLS certificate validation, which could allow an attacker to obtain sensitive information using man in the middle techniques.
-
CVE-2026-121818
Herowatch Mini developed by About Time Technologies has a system signing vulnerability. Authenticated local attackers can exploit a publicly available test key to sign malicious programs, allowing them to elevate privileges and execute arbitrary code.
-
CVE-2026-108014
Private keys for eight eMudhra intermediate CA certificates have been compromised https://emudhra.we-have-a-problem.com/ , which impacts certificate trust for outbound network request features of HCL Domino Server. An unauthenticated network attacker could spoof server certificates to execute a man in the middle (MITM) attack, enabling them to intercept and/or modify the content of outbound TLS connections.