Microsoft Internet Explorer 5.01, 5.5, and 6 allows remote attackers to cause a denial of service (application crash) and possibly execute arbitrary code via a web page with embedded CLSIDs that reference certain COM objects that are not intended for use within Internet Explorer, aka a variant of the "COM Object Instantiation Memory Corruption Vulnerability," a different vulnerability than CVE-2005-2127.

Microsoft Internet Explorer allows remote attackers to bypass cross-domain security restrictions and obtain sensitive information by using the @import directive to download files from other domains that are not valid Cascading Style Sheets (CSS) files, as demonstrated using Google Desktop, aka "CSSXSS" and "CSS Cross-Domain Information Disclosure Vulnerability."

The FTP client in Windows XP SP1 and Server 2003, and Internet Explorer 6 SP1 on Windows 2000 SP4, when "Enable Folder View for FTP Sites" is enabled and the user manually initiates a file transfer, allows user-assisted, remote FTP servers to overwrite files in arbitrary locations via crafted filenames.

Unknown vulnerability in Internet Explorer 5.0, 5.5, and 6.0 allows remote attackers to execute arbitrary code via a web site or an HTML e-mail containing a crafted JPEG image that causes memory corruption, aka "JPEG Image Rendering Memory Corruption Vulnerability".

Unknown vulnerability in Internet Explorer 5.0, 5.5, and 6.0 allows remote attackers to obtain information and possibly execute code when browsing from a web site to a web folder view using WebDAV, aka "Web Folder Behaviors Cross-Domain Vulnerability".

Internet Explorer 5.0, 5.5, and 6.0 allows remote attackers to cause a denial of service (application crash) and possibly execute arbitrary code via a web page with embedded CLSIDs that reference certain COM objects that are not ActiveX controls, including (1) devenum.dll, (2) diactfrm.dll, (3) wmm2filt.dll, (4) fsusd.dll, (5) dmdskmgr.dll, (6) browsewm.dll, (7) browseui.dll, (8) shell32.dll, (9) mshtml.dll, (10) inetcfg.dll, (11) infosoft.dll, (12) query.dll, (13) syncui.dll, (14) clbcatex.dll, (15) clbcatq.dll, (16) comsvcs.dll, and (17) msconf.dll, which causes memory corruption, aka "COM Object Instantiation Memory Corruption Vulnerability," a different vulnerability than CVE-2005-2087.

The JPEG decoder in Microsoft Internet Explorer allows remote attackers to cause a denial of service (CPU consumption or crash) and possibly execute arbitrary code via certain crafted JPEG images, as demonstrated using (1) mov_fencepost.jpg, (2) cmp_fencepost.jpg, (3) oom_dos.jpg, or (4) random.jpg.

Internet Explorer 5.01 SP4 up to 6 on various Windows operating systems, including IE 6.0.2900.2180 on Windows XP, allows remote attackers to cause a denial of service (application crash) and possibly execute arbitrary code via a web page with embedded CLSIDs that reference certain COM objects that are not ActiveX controls, as demonstrated using the JVIEW Profiler (Javaprxy.dll). NOTE: the researcher says that the vendor could not reproduce this problem.

Microsoft Internet Explorer 6 SP2 (6.0.2900.2180) crashes when the user attempts to add a URI to the restricted zone, in which the full domain name of the URI begins with numeric sequences similar to an IP address. NOTE: if there is not an exploit scenario in which an attacker can trigger this behavior, then perhaps this issue should not be included in CVE.

Internet Explorer 5.01, 5.5, and 6 allows remote attackers to execute arbitrary code via drag and drop events, aka the "Drag-and-Drop Vulnerability."

Internet Explorer 5.01, 5.5, and 6 allows remote attackers to spoof a less restrictive security zone and execute arbitrary code via an HTML page containing URLs that contain hostnames that have been double hex encoded, which are decoded twice to generate a malicious hostname, aka the "URL Decoding Zone Spoofing Vulnerability."

Internet Explorer 5.01, 5.5, and 6 does not properly validate buffers when handling certain DHTML methods including the createControlRange Javascript function, which allows remote attackers to execute arbitrary code, aka the "DHTML Method Heap Memory Corruption Vulnerability."

Internet Explorer 5.01, 5.5, and 6 does not properly validate certain URLs in Channel Definition Format (CDF) files, which allows remote attackers to obtain sensitive information or execute arbitrary code, aka the "Channel Definition Format (CDF) Cross Domain Vulnerability."

Internet Explorer 6.0 on Windows XP SP2 allows remote attackers to spoof the domain name of a URL in a titlebar for a script-initiated popup window, which could facilitate phishing attacks.

Race condition in the memory management routines in the DHTML object processor in Microsoft Internet Explorer 5.01, 5.5, and 6 allows remote attackers to execute arbitrary code via a malicious web page or HTML e-mail, aka "DHTML Object Memory Corruption Vulnerability".

Internet Explorer 6 on Windows XP SP2 allows remote attackers to bypass the file download warning dialog and possibly trick an unknowledgeable user into executing arbitrary code via a web page with a body element containing an onclick tag, as demonstrated using the createElement function.

Internet Explorer on Windows XP does not properly modify the "Drag and Drop or copy and paste files" setting when the user sets it to "Disable" or "Prompt," which may enable security-sensitive operations that are inconsistent with the user's intended configuration.

Internet Explorer 6.x on Windows XP SP2 allows remote attackers to execute arbitrary code, as demonstrated using a document with a draggable file type such as .xml, .doc, .py, .cdf, .css, .pdf, or .ppt, and using ADODB.Connection and ADODB.recordset to write to a .hta file that is interpreted in the Local Zone by HTML Help.

Heap-based buffer overflow in Internet Explorer 6 allows remote attackers to execute arbitrary code via long (1) SRC or (2) NAME attributes in IFRAME, FRAME, and EMBED elements, as originally discovered using the mangleme utility, aka "the IFRAME vulnerability" or the "HTML Elements Vulnerability."

Microsoft Internet Explorer 6.0 SP2 allows remote attackers to spoof a legitimate URL in the status bar and conduct a phishing attack via a web page that contains a BASE element that points to the legitimate site, followed by an anchor (a) element with an empty "href" attribute, and a FORM whose action points to a malicious URL, and an INPUT submit element that is modified to look like a legitimate URL.