IBM API Connect 10.0.8.0 through 10.0.8.12 and 12.1.0.0 through 12.1.1.2 could allow a token to be accepted that should be rejected due to an improperly implemented security check. Under certain conditions, token validation is incomplete, so a token may be accepted for a purpose it was not intended for.
CVE-2026-96303 is a vulnerability in Microsoft PowerShell 7.4 involving an improperly implemented security check that allows an unauthorized attacker to bypass a security feature over a network. The issue affects versions 7.4.0 through 7.4.20, 7.5.0 through 7.5.11, and 7.6.0 through 7.6.7. A fix has been officially released in versions 7.4.21, 7.5.12, and 7.6.8. The vulnerability has a CVSS 3.1 base score of 6.5, indicating medium severity. There are no known exploits in the wild at this time.
CVE-2026-70174 is a high-severity vulnerability in Microsoft .NET 10.0 runtime caused by an improperly implemented security check. This flaw allows an unauthorized attacker to execute code remotely over a network. Multiple versions of the .NET runtime are affected, with official fixes released in versions 10.0.13 and later. The vulnerability has a CVSS score of 8.8, indicating a significant risk to confidentiality, integrity, and availability.
CVE-2026-85805 is a vulnerability in Windows PowerShell on Microsoft Windows 10 Version 1607 and other specified versions. It involves an improperly implemented security check that allows an unauthorized attacker to bypass a security feature over a network. The vulnerability has a medium severity score of 6.5 and has been officially fixed in multiple Windows 10 builds.
Authlib version 1.8.0 and earlier contains a critical vulnerability in its JsonWebSignature.deserialize_json() method, which bypasses signature verification. This flaw allows the method to return the payload as successfully verified without actually checking the signature or requiring a cryptographic key.
In NLnet Labs Unbound 1.6.0 up to and including 1.25.1, a replay of a wildcard rrset as another piece of data, could be briefly considered DNSSEC secure based only on the RRSIG validation and stored into cache, before later validation treats it as bogus based on NSEC validation. When the resolving thread puts secure on the rrset, and another thread that is on the serve expired path then picks up the updated rrset contents with the secure status for a reply, it can be used to change a specific record, next to a wildcard that could be covered by the wildcard, into the wildcard. A malicious actor can exploit the possible poisonous effect by having any DNSSEC-singed domain (irrelevant to the victim domain) and a CNAME wrapper record that points to a record next to a wildcard (that could be covered by the wildcard). Then quering Unbound for the wildcard sibling record would seed the secure message. A later (after expiry) query for the CNAME wrapper would need to resolve the target sibling record. If the wildcard replay is injected into the response, the wildcard rrset will update the expired sibling record with a secure status before completing proper wildcard validation with NSEC records and eventually treating the CNAME wrapper answer as bogus. The updated poisoned rrset is now secure and points to the wildcard. This vulnerability is explicit for the serve expired path and needs injection of the signed wildcard rrset without the NSEC accompanying rrset.
Trezor Safe 3, Safe 5, and Safe 7 firmware contains a confirmation-binding flaw in the Ethereum sign_tx / sign_tx_eip1559 flow. For contract interactions, the device confirms only the initial calldata chunk while the signature commits to the full streamed calldata. An attacker could present calldata to a victim then supply a different tail that changes the signed transaction. Fixed in 70c9b0c.
In liboauth2 the Demonstrating Proof-of-Possession (DPoP) verifier accepts a proof whose JSON Web Key (jwk) header contains private key material. RFC 9449 section 4.3 step 7 requires the verifier to reject such a proof but oauth2_token_verify() function returns success for a malformed DPoP proof that embeds the private Elliptic Curve (EC) key in the header.
This issue was fixed in version 2.3.0
It is possible to bypass the Kerberos pre-authentication check in Apache Kerby by sending a PA-DATA with an unrecognized or unsupported type. Users are recommended to upgrade to version 2.1.2, which fixes this issue.
Backpropagate is a Python library for fine-tuning large language models on a single GPU. In versions 1.1.0 and 1.1.1, the optional Reflex web UI exposes a training control plane without authentication: dataset upload, model load, training start/stop, multi-run orchestration, GGUF export, and HuggingFace Hub push. The CLI accepts two operator-facing flags intended as security controls: --auth user:pass — documented as "require HTTP Basic authentication on every request to the UI." and--share — documented as "expose the UI on a public address; requires --auth." When --auth user:pass is passed, the CLI prints Auth: enabled (user: <username>) to confirm to the operator that authentication is active, then exports BACKPROPAGATE_UI_AUTH=user:pass to the subprocess that launches the Reflex backend. The Reflex backend (backpropagate/ui_app/**) never reads BACKPROPAGATE_UI_AUTH. No authentication middleware is registered. No request-level guard runs. No WebSocket upgrade guard runs. Any client that reaches the bound port — local or remote, depending on whether --share is used — has full UI access. An inline comment at backpropagate/cli.py:1217-1218 in the v1.1.0 source documents the gap: "For Phase 1 the variable is exported but Reflex doesn't read it yet." This comment was internal-facing; the user-facing documentation (README, CHANGELOG, SHIP_GATE) advertised the contract as enforced. An attacker who reaches the bound port can read uploaded datasets, trigger arbitrary training runs against any local base models as well as read their paths, trigger HuggingFace Hub pushes and cause disk-fill DoS. This issue has been fixed in version 1.2.0. If developers cannot immediately upgrade to 1.2.0 run backprop ui with no flags so it binds to localhost, use SSH port-forwarding (ssh -L 7860:localhost:7860 <training-host>) instead of --share for remote access, and audit any host previously launched with --share, re-issuing any HF tokens used during those sessions.
A logic error in OAuthRequestFilter rejects legitimate requests originating from the bound IP address, while blindly allowing requests from any other IP address. Enabling this
security feature inadvertently creates an inverse security check. Users are recommended to upgrade to versions 4.2.2 or 4.1.7 or 3.6.12, which fixes this issue.
CVE-2026-44473 is a high-severity vulnerability in ellanetworks core (Ella Core), a 5G core designed for private networks. Versions prior to 1.10.0 allow a radio with a valid NG Setup to send a forged PDUSessionResourceSetupResponse containing any UE's AMF-UE-NGAP-ID. The core fails to verify that the message arrived on the SCTP association bound to that UE's logical NG-connection, resulting in the creation of a GTP tunnel towards the radio. This improper security check can lead to denial of service or disruption of network operations. The vulnerability is fixed in version 1.10.0. No known exploits in the wild have been reported.
Ella Core versions prior to 1.10.0 contain a vulnerability where the UE Security Capabilities received in NGAP PathSwitchRequest messages are not verified against locally stored values. This allows a malicious gNB to overwrite stored UE security capabilities with arbitrary values by sending a crafted PathSwitchRequest message. The vulnerability is fixed in version 1.10.0. The CVSS score is 6.1, indicating a medium severity issue with potential integrity and availability impacts but no confidentiality loss.
Ella Core is a 5G core designed for private networks. Prior to 1.10.0, Ella Core didn't enforce security rules on concurrent running of security procedures defined in TS 33.501 §6.9.5.1 — it could send a NAS Security Mode Command while an N2 handover was still pending (and vice versa). Concurrent Security Mode Command and N2 handover produce a KgNB mismatch between the UE and target gNB, causing the handover to fail. Requires a stalled gNB + re-registration race to trigger. This vulnerability is fixed in 1.10.0.
The in-memory keyring returned by NewKeyring() silently accepted keys with the ConfirmBeforeUse constraint but never enforced it. The key would sign without any confirmation prompt, with no indication to the caller that the constraint was not in effect. NewKeyring() now returns an error when unsupported constraints are requested.
Next.js is a React framework for building full-stack web applications. From 15.2.0 to before 15.5.18 and 16.2.6, it was found that the fix addressing CVE-2026-44575 did not apply to middleware.ts with Turbopack. This vulnerability is fixed in 15.5.18 and 16.2.6.
HCL BigFix Service Management (SM) is affected by a security misconfiguration vulnerability due to CSP header. This could allow attackers to inject malicious scripts increasing the risk of cross-site scripting (XSS) and potential exposure of sensitive information.
HCL DFXAnalytics is affected by an Insecure Security Header configuration vulnerability where the Content-Security-Policy does not define strict directives for object-src and base-uri, which could allow an attacker to exploit injection vectors such as Cross-Site Scripting (XSS)