Parsing documents as HTML in Downloads in Google Chrome prior to 66.0.3359.117 allowed a remote attacker to cause Chrome to execute scripts via a local non-HTML page.

Insufficient policy enforcement in Extensions API in Google Chrome prior to 67.0.3396.62 allowed an attacker who convinced a user to install a malicious extension to bypass navigation restrictions via a crafted Chrome Extension.

The ssl_Do1stHandshake function in sslsecur.c in libssl in Mozilla Network Security Services (NSS) before 3.15.4, when the TLS False Start feature is enabled, allows man-in-the-middle attackers to spoof SSL servers by using an arbitrary X.509 certificate during certain handshake traffic.

The child frames in Mozilla Firefox before 1.5.0.10 and 2.x before 2.0.0.2, and SeaMonkey before 1.0.8 inherit the default charset from the parent window, which allows remote attackers to conduct cross-site scripting (XSS) attacks, as demonstrated using the UTF-7 character set.

Insufficient policy enforcement in extensions API in Google Chrome prior to 69.0.3497.81 allowed an attacker who convinced a user to install a malicious extension to bypass navigation restrictions via a crafted Chrome Extension.

A vulnerability in register allocation in JavaScript can lead to type confusion, allowing for an arbitrary read and write. This leads to remote code execution inside the sandboxed content process when triggered. This vulnerability affects Firefox ESR < 60.2.2 and Firefox < 62.0.3.

WebExtensions could use the "mozAddonManager" API by modifying the CSP headers on sites with the appropriate permissions and then using host requests to redirect script loads to a malicious site. This allows a malicious extension to then install additional extensions without explicit user permission. This vulnerability affects Firefox < 51.

The Tencent WBlog (com.tencent.WBlog) 3.3.1 and MicroBlogPad 1.4.0 applications for Android do not properly protect data, which allows remote attackers to read or modify message drafts and search keywords via a crafted application.

The 'Copy Image Link' context menu action would copy the final image URL after redirects. By embedding an image that triggered authentication flows - in conjunction with a Content Security Policy that stopped a redirection chain in the middle - the final image URL could be one that contained an authentication token used to takeover a user account. If a website tricked a user into copy and pasting the image link back to the page, the page would be able to steal the authentication tokens. This was fixed by making the action return the original URL, before any redirects. This vulnerability affects Firefox < 94.

Mozilla Firefox before 20.0 and SeaMonkey before 2.17 do not prevent origin spoofing of tab-modal dialogs, which allows remote attackers to conduct phishing attacks via a crafted web site.

Using techniques that built on the slipstream research, a malicious webpage could have exposed both an internal network's hosts as well as services running on the user's local machine. This vulnerability affects Firefox < 84, Thunderbird < 78.6, and Firefox ESR < 78.6.

When a user typed a URL in the address bar or the search bar and quickly hit the enter key, a website could sometimes capture that event and then redirect the user before navigation occurred to the desired, entered address. To construct a convincing spoof the attacker would have had to guess what the user was typing, perhaps by suggesting it. This vulnerability affects Firefox < 84.

Mozilla Firefox before 18.0 on Android and SeaMonkey before 2.15 do not restrict a touch event to a single IFRAME element, which allows remote attackers to obtain sensitive information or possibly conduct cross-site scripting (XSS) attacks via a crafted HTML document.

The contribution feature in Zamboni does not verify that the server hostname matches a domain name in the subject's Common Name (CN) or subjectAltName field of the X.509 certificate, which allows man-in-the-middle attackers to spoof SSL servers via an arbitrary valid certificate, related to use of the Python urllib2 library.

The developer-account sample code in Google AdMob does not verify that the server hostname matches a domain name in the subject's Common Name (CN) or subjectAltName field of the X.509 certificate, which allows man-in-the-middle attackers to spoof SSL servers via an arbitrary valid certificate.

Yahoo! Toolbar 1.0.0.5 and earlier for Chrome and Safari allows remote attackers to modify the configured search URL, and intercept search terms, via a crafted web page.

The Tencent QQPimSecure (com.tencent.qqpimsecure) application 3.0.2 for Android does not properly protect data, which allows remote attackers to read or modify SMS/MMS messages and a contact list via a crafted application.

google-checkout-php-sample-code before 1.3.2 does not verify that the server hostname matches a domain name in the subject's Common Name (CN) or subjectAltName field of the X.509 certificate, which allows man-in-the-middle attackers to spoof SSL servers via an arbitrary valid certificate.

By exploiting an Open Redirect vulnerability on a website, an attacker could have spoofed the site displayed in the download file dialog to show the original site (the one suffering from the open redirect) rather than the site the file was actually downloaded from. This vulnerability affects Firefox < 81, Thunderbird < 78.3, and Firefox ESR < 78.3.

The Tencent MobileQQ (com.tencent.mobileqq) application 2.2 for Android does not properly protect data, which allows remote attackers to read or modify messages and a friends list via a crafted application.

In Tensorflow before version 2.3.1, the `SparseCountSparseOutput` implementation does not validate that the input arguments form a valid sparse tensor. In particular, there is no validation that the `indices` tensor has the same shape as the `values` one. The values in these tensors are always accessed in parallel. Thus, a shape mismatch can result in accesses outside the bounds of heap allocated buffers. The issue is patched in commit 3cbb917b4714766030b28eba9fb41bb97ce9ee02 and is released in TensorFlow version 2.3.1.

Google Chrome before 17.0.963.46 does not properly implement the drag-and-drop feature, which makes it easier for remote attackers to spoof the URL bar via unspecified vectors.

In TensorFlow Lite before versions 1.15.4, 2.0.3, 2.1.2, 2.2.1 and 2.3.1, saved models in the flatbuffer format use a double indexing scheme: a model has a set of subgraphs, each subgraph has a set of operators and each operator has a set of input/output tensors. The flatbuffer format uses indices for the tensors, indexing into an array of tensors that is owned by the subgraph. This results in a pattern of double array indexing when trying to get the data of each tensor. However, some operators can have some tensors be optional. To handle this scenario, the flatbuffer model uses a negative `-1` value as index for these tensors. This results in special casing during validation at model loading time. Unfortunately, this means that the `-1` index is a valid tensor index for any operator, including those that don't expect optional inputs and including for output tensors. Thus, this allows writing and reading from outside the bounds of heap allocated arrays, although only at a specific offset from the start of these arrays. This results in both read and write gadgets, albeit very limited in scope. The issue is patched in several commits (46d5b0852, 00302787b7, e11f5558, cd31fd0ce, 1970c21, and fff2c83), and is released in TensorFlow versions 1.15.4, 2.0.3, 2.1.2, 2.2.1, or 2.3.1. A potential workaround would be to add a custom `Verifier` to the model loading code to ensure that only operators which accept optional inputs use the `-1` special value and only for the tensors that they expect to be optional. Since this allow-list type approach is erro-prone, we advise upgrading to the patched code.

Google Chrome before 18.0.1025.142 does not properly check X.509 certificates before use of a SPDY proxy, which might allow man-in-the-middle attackers to spoof servers or obtain sensitive information via a crafted certificate.

Google Chrome before 11.0.696.57 allows remote attackers to spoof the URL bar via vectors involving (1) a navigation error or (2) an interrupted load.

Google Chrome before 11.0.696.57 allows user-assisted remote attackers to spoof the URL bar via vectors involving a redirect and a manual reload.

The Math.random function in the JavaScript implementation in Mozilla Firefox 3.5.10 through 3.5.11, 3.6.4 through 3.6.8, and 4.0 Beta1 uses a random number generator that is seeded only once per document object, which makes it easier for remote attackers to track a user, or trick a user into acting upon a spoofed pop-up message, by calculating the seed value, related to a "temporary footprint" and an "in-session phishing attack." NOTE: this vulnerability exists because of an incorrect fix for CVE-2008-5913.

The js_InitRandom function in the JavaScript implementation in Mozilla Firefox 3.5.x before 3.5.10 and 3.6.x before 3.6.4, and SeaMonkey before 2.0.5, uses the current time for seeding of a random number generator, which makes it easier for remote attackers to guess the seed value via a brute-force attack, a different vulnerability than CVE-2008-5913.

The js_InitRandom function in the JavaScript implementation in Mozilla Firefox 3.5.10 through 3.5.11, 3.6.4 through 3.6.8, and 4.0 Beta1 uses a context pointer in conjunction with its successor pointer for seeding of a random number generator, which makes it easier for remote attackers to guess the seed value via a brute-force attack, a different vulnerability than CVE-2010-3171.

Mozilla Firefox before 3.5.14 and 3.6.x before 3.6.11, Thunderbird before 3.0.9 and 3.1.x before 3.1.5, and SeaMonkey before 2.0.9 do not properly handle certain modal calls made by javascript: URLs in circumstances related to opening a new window and performing cross-domain navigation, which allows remote attackers to bypass the Same Origin Policy via a crafted HTML document.

src/net/http/http_transaction_winhttp.cc in Google Chrome before 1.0.154.53 uses the HTTP Host header to determine the context of a document provided in a (1) 4xx or (2) 5xx CONNECT response from a proxy server, which allows man-in-the-middle attackers to execute arbitrary web script by modifying this CONNECT response, aka an "SSL tampering" attack.

Insufficient policy enforcement in service workers in Google Chrome prior to 74.0.3729.108 allowed a remote attacker to bypass navigation restrictions via a crafted HTML page.

A Content-Security-Policy that blocks in-line scripts could be bypassed using an object tag to execute JavaScript in the protected document (cross-site scripting). This is a separate bypass from CVE-2019-17000.*Note: This flaw only affected Firefox 69 and was not present in earlier versions.*. This vulnerability affects Firefox < 70.

An object tag with a data URI did not correctly inherit the document's Content Security Policy. This allowed a CSP bypass in a cross-origin frame if the document's policy explicitly allowed data: URIs. This vulnerability affects Firefox < 70.

The JavaScript implementation in Mozilla Firefox 3.x before 3.5.10 and 3.6.x before 3.6.4, and SeaMonkey before 2.0.5, allows remote attackers to send selected keystrokes to a form field in a hidden frame, instead of the intended form field in a visible frame, via certain calls to the focus method.

Mozilla Firefox before 48.0 does not properly restrict drag-and-drop (aka dataTransfer) actions for file: URIs, which allows user-assisted remote attackers to access local files via a crafted web site.

Adobe Flash Player before 18.0.0.343 and 19.x through 21.x before 21.0.0.213 on Windows and OS X and before 11.2.202.616 on Linux allows attackers to bypass intended access restrictions via unspecified vectors.

In all Android releases from CAF using the Linux kernel, HLOS can overwite secure memory or read contents of the keystore.

Adobe Flash Player before 18.0.0.343 and 19.x through 21.x before 21.0.0.213 on Windows and OS X and before 11.2.202.616 on Linux allows attackers to bypass the ASLR protection mechanism via JIT data.

An issue was discovered on Samsung mobile devices with KK(4.4) and later software through 2015-06-16. In some cases, HTTP is used for an Inputmethod, rather than HTTPS. A man-in-the-middle attacker can modify the client-server data stream to insert directory traversal sequences into an extracted file path. The Samsung ID is SVE-2015-4363 (November 2015).

Mozilla Firefox before 1.5.0.5, Thunderbird before 1.5.0.5, and SeaMonkey before 1.0.3 allows remote attackers to hijack native DOM methods from objects in another domain and conduct cross-site scripting (XSS) attacks using DOM methods of the top-level object.

Insufficient validation of untrusted input in Intents in Google Chrome on Android prior to 95.0.4638.69 allowed a remote attacker to arbitrarily browser to a malicious URL via a crafted HTML page.

Inappropriate implementation in Navigation in Google Chrome on Windows prior to 94.0.4606.54 allowed a remote attacker to inject scripts or HTML into a privileged page via a crafted HTML page.

The Upgrade-Insecure-Requests (UIR) specification states that if UIR is enabled through Content Security Policy (CSP), navigation to a same-origin URL must be upgraded to HTTPS. Firefox will incorrectly navigate to an HTTP URL rather than perform the security upgrade requested by the CSP in some circumstances, allowing for potential man-in-the-middle attacks on the linked resources. This vulnerability affects Firefox < 66.

If two same-origin documents set document.domain differently to become cross-origin, it was possible for them to call arbitrary DOM methods/getters/setters on the now-cross-origin window. This vulnerability affects Firefox < 70, Thunderbird < 68.2, and Firefox ESR < 68.2.

Application permissions give additional remote troubleshooting permission to the site input.mozilla.org, which has been retired and now redirects to another site. This additional permission is unnecessary and is a potential vector for malicious attacks. This vulnerability affects Firefox < 68.

By using a form with a data URI it was possible to gain access to the privileged JSONView object that had been cloned into content. Impact from exposing this object appears to be minimal, however it was a bypass of existing defense in depth mechanisms. This vulnerability affects Firefox < 70, Thunderbird < 68.2, and Firefox ESR < 68.2.

btif/src/btif_dm.c in Android before 5.1 does not properly enforce the temporary nature of a Bluetooth pairing, which allows user-assisted remote attackers to bypass intended access restrictions via crafted Bluetooth packets after the tapping of a crafted NFC tag.

Insufficient policy enforcement in content security policy in Google Chrome prior to 91.0.4472.77 allowed a remote attacker to bypass content security policy via a crafted HTML page.

Firefox incorrectly accepted a newline in a HTTP/3 header, interpretting it as two separate headers. This allowed for a header splitting attack against servers using HTTP/3. This vulnerability affects Firefox < 91.0.1 and Thunderbird < 91.0.1.