A malicious image can advertise DiffIDs from another image while containing different layer contents. In affected versions, BuildKit could use the advertised DiffIDs to derive cache and snapshot identity without validating that they matched the actual layer contents. If a BuildKit daemon with shared or persistent cache first processes such a malicious image, a later build using the victim image may mount the attacker-controlled layer contents as the base image. This can allow code from the malicious image to run in the victim build, for example by replacing a commonly executed path such as /bin/sh. The attacker-controlled code may read build secrets mounted into the build, access other build resources, alter output artifacts, or hang the build. The issue affects both regular snapshotters and lazy-pulling snapshotters such as stargz.
CVE-2026-71888 is a high-severity vulnerability in Bouncy Castle for Java before version 1.86 and certain LTS and FIPS releases. The issue arises from improper validation of the integrity check value in the streaming CMS AuthenticatedData parser, where a mismatch between digestAlgorithm and authAttrs fields could allow an attacker to insert unauthenticated attributes into a message. This could lead to applications acting on attacker-controlled attributes without the attacker having the encryption or MAC keys. The vulnerability has been fixed in versions 1.86, 2.73.13, 2.1.13, 2.0.13, 1.0.13, 2.1.8, and 2.0.8.
CVE-2026-85515 is a vulnerability in Bouncy Castle for Java before version 1.86 and certain LTS and FIPS editions that causes improper validation of integrity check values in OpenPGP encrypted messages. Specifically, truncated encrypted messages may be accepted without error, and on some code paths, no integrity check is performed at all, allowing altered ciphertext to be accepted as valid plaintext. This affects the SEIPD version 1 path more severely, where no integrity verification occurs, and the AEAD path where truncated chunks may be silently dropped without error. The issue has been fixed in versions 1.86, 2.73.13, 1.0.14, 2.0.14.1, and 2.1.14.
Release of unverified plaintext in the CCM (CcmBlockCipher) and DSTU 7624 CCM (KCcmBlockCipher) AEAD modes in Legion of the Bouncy Castle Inc. bc-csharp before 2.7.0 allows a remote attacker to obtain decryptions of ciphertexts of their choosing via forged messages sent to an application that lets the output buffer of a failed decryption be observed, for example through buffer reuse or logging, because decryption wrote the recovered plaintext into the caller-supplied output buffer before checking the authentication tag and left it there when the check failed. Only decryption into a caller-supplied buffer is affected; methods that return a newly allocated array are not.
Exposure of the message authentication key through the encryption keystream in the stream mode of IesEngine (an IesEngine constructed without a block cipher) in Legion of the Bouncy Castle Inc. bc-csharp before 2.7.0 allows a remote attacker who has observed one encrypted message with known plaintext to forge shorter messages of their choosing that the recipient accepts as authentic, via a crafted ciphertext and MAC tag, because the MAC key was taken from the key derivation output directly after a keystream as long as the message, while the derivation input depends only on the static key pair and fixed parameters. The keystream revealed by that one message therefore contains the MAC key for every sufficiently shorter message.
Improper validation of integrity check value in the AES-CCM implementation (CcmParameters and CcmBlockCipher) in Legion of the Bouncy Castle Inc. bc-csharp before 2.7.0 allows an on-path attacker to modify CCM-encrypted content without detection via an AlgorithmIdentifier whose CCMParameters declare an authentication tag (aes-ICVlen) of zero or another length outside the RFC 5084 set, because CcmParameters accepted any value and CcmBlockCipher validated the tag length only when encrypting, so decryption compared a zero-length or very short tag. Affected paths include ParameterUtilities.GetCipherParameters, used by CmsEnvelopedData and CmsEnvelopedDataParser for EnvelopedData encrypted with AES-CCM, and any caller passing an unchecked tag length to CcmBlockCipher for decryption.
CVE-2026-102759 is a medium severity vulnerability in Eclipse Foundation's NetX Duo where the NetX Secure TLS component improperly validates the integrity check value. Specifically, it accepts an empty TLS application-data record without verifying its message authentication code (MAC), treating such records as valid and advancing the receive sequence number without MAC comparison. Empty TLS application-data records are legal and commonly used in TLS 1.0 as a mitigation against the BEAST attack.
CVE-2026-95625 is a vulnerability in the Tauri tauri-plugin-updater where the update manifest is not signed or authenticated, allowing an attacker to serve a crafted manifest. This can force the installation of older signed releases despite the anti-rollback check, which only compares an unsigned version field. The vulnerability affects versions >=2.0.0 and later. The CVSS score is 5.9, indicating medium severity.
CVE-2026-92701 is a critical origin validation vulnerability in Cocos AI versions 0.8.2 and earlier. The Intel TDX aTLS verifier's VerifTeeAttestation function fails to apply the expected current-session value to TdQuoteBodyPolicy.ReportData before validating TDX QuoteV4 evidence. This allows acceptance of structurally valid but mismatched or reused REPORT_DATA, potentially causing trust or authorization decisions to be made for unrelated attestation contexts. The issue was fixed in version 0.9.0 by removing the affected verifier path during migration to post-handshake attestation.
CVE-2026-54578 is a vulnerability in the MidnightBSD package manager 'mport' prior to version 2.7.8. The function mport_verify_package() could continue processing after failures in MD5 or SHA256 checksum computations, leading to comparisons against stale hash data instead of freshly computed digests. This flaw allows an attacker who can influence an installed file or cause hashing failures to potentially receive misleading integrity verification results or conceal checksum failures. The issue is resolved in version 2.7.8.
mport is the MidnightBSD Package Manager. Prior to 2.7.8, libmport/util.c did not make every truncated, corrupt, or failed zstd stream fatal in mport_decompress_zstd(), and libmport/fetch.c did not consistently propagate those failures to index-fetch callers. A malicious or faulty mirror could supply compressed package index data that caused ZSTD_decompressStream() or an output write to fail while leaving partial index output available for later use, resulting in package-index integrity loss or denial of service. This issue is fixed in version 2.7.8.
CVE-2026-73459 is a high-severity vulnerability in Arista Networks EOS affecting versions 4.36.0 through 4.36.1F. It involves improper validation of integrity check values in the IS-IS protocol implementation. An unauthenticated attacker able to inject specially crafted IS-IS LSP PDUs can cause legitimate LSPs to be purged unexpectedly from the IS-IS link-state database, potentially resulting in traffic loss.
CVE-2026-54174 is a vulnerability in chainguard-dev's melange and apko tools used for building apk packages. Versions of apko prior to 1.2.9 and melange prior to 0.50.4 verified the control section hash against the signed APKINDEX but did not verify the data section hash of the actual package files. This flaw allows an attacker who can compromise a mirror, cache, or perform a man-in-the-middle attack to substitute arbitrary file contents while passing the control hash check. The issue is fixed in apko 1.2.9 and melange 0.50.4.
CVE-2026-72929 is a high-severity vulnerability in Microsoft Windows 11 version 23H2 involving improper validation of an integrity check value in Windows Installer. This flaw allows an authorized local attacker to elevate privileges. The issue affects multiple builds of Windows 11 23H2 prior to specific fixed versions. Microsoft has released official fixes for all affected versions.
In Bouncy Castle for Java before 1.85, OpenPGP AEAD decryption skips final tag on chunk-aligned data. This issue also affects Bouncy Castle for Java LTS before 2.73.12, and Bouncy Castle for Java FIPS (BC-FJA) before bcpg-fips 1.0.13 (1.0.X series), 2.0.13 (2.0.X series) and 2.1.13 (2.1.X series).
Bouncy Castle for Java versions before 1.85 and LTS versions before 2.73.12 have a vulnerability in the KCCMBlockCipher MAC implementation where the nonce is not bound when Additional Authenticated Data (AAD) is absent. This flaw allows a cross-nonce AEAD forgery attack, undermining the integrity guarantees of the cipher.
A vulnerability exists in Bouncy Castle for Java before version 1.85 and in the LTS version before 2.73.12, where the IESEngine stream-mode MAC can be forged due to a length-dependent key derivation function (KDF) split. This issue is tracked as CVE-2026-12816 and is classified under CWE-354 (Improper Validation of Integrity Check Value).
Bouncy Castle for Java versions prior to 1.85, including certain LTS and FIPS releases, have a vulnerability in CCM-family encryption modes where plaintext is written to the caller's buffer before the authentication tag is verified. This behavior can lead to potential exposure of plaintext data if the tag check fails.