News & Sicurezza
Aggiornamenti da ENISA, NVD e le principali fonti di cybersecurity europee. Tutto quello che un Responsabile Tecnico deve sapere.
55853 risultati
CVE ID :CVE-2026-41262 Published : Aug. 26, 2026, 7:16 p.m. | 51 minutes ago Description :Fleet is an open-source device management platform built on osquery. In versions prior to 4.85.0, the global policy read endpoint (GET /api/latest/fleet/policies/{policy_id}) fails to verify team ownership of the requested policy, allowing an authenticated user with observer-level access on any single team to read the full details of policies belonging to any other team and bypass Fleet's team isolation model. The handler authorizes the request against an empty policy object whose TeamID is nil, which an authorization rule permits for any user holding a role on any team, and then fetches the policy by ID with no team filter and returns it without any post-fetch scope check. Because policy IDs are sequential integers, an attacker can enumerate them to read other teams' policy SQL queries, host pass and fail counts, and associated software-installer and script metadata, exposing security-monitoring strategies and compliance posture across team boundaries. This issue is fixed in version 4.85.0. Severity: 4.3 | MEDIUM Visit the link for more details, such as CVSS details, affected products, timeline, and more...
CVE ID :CVE-2026-48786 Published : Aug. 26, 2026, 7:16 p.m. | 51 minutes ago Description :Fleet is an open-source device management platform built on osquery. In versions prior to 4.87.0, the target search endpoint (POST /api/latest/fleet/targets) returned unmasked team enroll secrets and full team configuration, including credential-bearing agent options, to low-privilege observer-class users. Other team-facing endpoints mask these fields for observers, but the target search endpoint did not apply the same sanitization, so an authenticated user with the Observer, Observer+, or Technician role, whether global or team-scoped, could retrieve the secrets and agent options by performing a target search against an observer-runnable query. With a leaked team enroll secret an attacker could enroll unauthorized hosts into the affected team, and if the team's agent options contained credentials such as AWS secret access keys or proxy passwords, those values were disclosed as well. This issue is fixed in version 4.87.0. Severity: 6.5 | MEDIUM Visit the link for more details, such as CVSS details, affected products, timeline, and more...
CVE ID :CVE-2026-19485 Published : Aug. 26, 2026, 7:16 p.m. | 51 minutes ago Description :A Predictable Resource Name vulnerability in BigQuery Import Staging in Google Cloud Vertex AI Search for Commerce versions prior to 2026-04-27 on Google Cloud Platform allows an attacker knowing the victim's project number to obtain read/write access to staged data and error logs using predictable bucket names. This vulnerability was patched and no customer action is needed. Severity: 9.3 | CRITICAL Visit the link for more details, such as CVSS details, affected products, timeline, and more...
CVE ID :CVE-2026-36851 Published : Aug. 26, 2026, 7:16 p.m. | 51 minutes ago Description :Path traversal vulnerability in UnPoller 2.33.0 password field allows arbitrary file read and network exfiltration. Severity: 0.0 | NA Visit the link for more details, such as CVSS details, affected products, timeline, and more...
CVE ID :CVE-2025-61165 Published : Aug. 26, 2026, 7:16 p.m. | 51 minutes ago Description :An arbitrary file upload vulnerability in the /v1/my_drive/batch_upload component of cohere North AI v1.1.5 allows attackers to exeute arbitrary code via uploading a crafted file. Severity: 0.0 | NA Visit the link for more details, such as CVSS details, affected products, timeline, and more...
CVE ID :CVE-2025-61164 Published : Aug. 26, 2026, 7:16 p.m. | 51 minutes ago Description :Cohere North AI v1.1.5 was discovered to contain an information leak via the WebSocket Endpoint. Severity: 0.0 | NA Visit the link for more details, such as CVSS details, affected products, timeline, and more...
CVE ID :CVE-2025-61163 Published : Aug. 26, 2026, 7:16 p.m. | 51 minutes ago Description :Cohere North AI v1.1.5 was discovered to contain excessively permissive cross-domain policy with untrusted domains. This occurs via the server failing to validate the Origin header of incoming connection requests. Severity: 0.0 | NA Visit the link for more details, such as CVSS details, affected products, timeline, and more...
CVE ID :CVE-2025-61162 Published : Aug. 26, 2026, 7:16 p.m. | 51 minutes ago Description :Incorrect access control in Cohere North AI v1.1.5 allows attackers to arbitrarily overwrite user info via a crafted request to the /api/internal/v1/users/{{USER_ID}} endpoint Severity: 0.0 | NA Visit the link for more details, such as CVSS details, affected products, timeline, and more...
CVE ID :CVE-2026-54245 Published : Aug. 26, 2026, 7:13 p.m. | 54 minutes ago Description :Fleet is an open-source device management platform built on osquery. In versions prior to 4.86.2, the Okta conditional access integration in Fleet Premium is vulnerable to SQL injection through a host-supplied value that is used in a database query without proper parameterization, allowing an attacker who controls a single enrolled host to read or modify arbitrary data in the Fleet database. The value is reported by the host's own agent and stored verbatim, then used on an unauthenticated request path that supports the conditional access integration, so any party controlling one enrolled host, the lowest-privilege position in the product, can influence the query. By disclosing arbitrary database contents an attacker can extract stored session tokens and replay them to act as a global administrator, and on a managed fleet that administrator access enables running scripts on enrolled hosts, leading to remote code execution. The issue requires Fleet Premium with the Okta conditional access integration enabled and does not affect instances where it is not configured. This issue is fixed in version 4.86.2. Severity: 0.0 | NA Visit the link for more details, such as CVSS details, affected products, timeline, and more...
CVE ID :CVE-2026-74771 Published : Aug. 26, 2026, 7:11 p.m. | 56 minutes ago Description :Dell PowerProtect One, versions 20.1.0.0 and below, contain an Authorization Bypass Through User-Controlled Key vulnerability. A low privileged attacker with remote access could potentially exploit this vulnerability, leading to Information tampering. Severity: 6.5 | MEDIUM Visit the link for more details, such as CVSS details, affected products, timeline, and more...
CVE ID :CVE-2026-46371 Published : Aug. 26, 2026, 7:06 p.m. | 1 hour, 2 minutes ago Description :Fleet is an open-source device management platform built on osquery. In versions up to and including 4.84.1, the Apple MDM commands listing endpoint (GET /api/v1/fleet/mdm/apple/commands) allowed an authenticated user with the lowest-privilege Observer role to extract sensitive values from joined database tables, including host enrollment secrets and Apple Push Notification Service tokens, through a sort-order oracle. The endpoint accepted a user-supplied order_key parameter that was not validated against a column allowlist, and because the underlying query joins the hosts and nano_enrollments tables, an attacker could set the sort column to a sensitive field and combine it with the cursor-based after parameter to binary-search the value one character at a time, with the presence or absence of results revealing each character even though the value never appeared in the response. With extracted node_key or orbit_node_key values an attacker could impersonate enrolled hosts to Fleet's osquery and Orbit endpoints, submit fabricated host data, and retrieve pending scripts and commands. This issue is fixed in version 4.84.2. Severity: 0.0 | NA Visit the link for more details, such as CVSS details, affected products, timeline, and more...
whichllm before 0.5.16 contains a code injection vulnerability in the run and snippet commands that allows a remote attacker who controls a HuggingFace repository to achieve arbitrary code execution by crafting a malicious GGUF filename containing double quotes or other special characters. The script generation function in cli.py interpolates HuggingFace-derived values, including GGUF variant filenames from the Hub API siblings rfilename field, directly into Python source code without escaping, allowing the crafted filename to break out of the generated string literal and execute injected code on the user's machine before any model download occurs.
Pagina 1641 di 4655