Cerulean Studios Trillian 3.1 Basic does not check SSL certificates during MSN authentication, which allows remote attackers to obtain MSN credentials via a man-in-the-middle attack with a spoofed SSL certificate.

Buffer overflow in the XML parser in Trillian 3.1.9.0, and possibly earlier, allows remote attackers to cause a denial of service (crash) and possibly execute arbitrary code via a crafted DTD file.

Double free vulnerability in the XML parser in Trillian before 3.1.12.0 allows remote attackers to execute arbitrary code via a crafted XML expression, related to the "IMG SRC ID."

Heap-based buffer overflow in the XML parser in the AIM plugin in Trillian before 3.1.12.0 allows remote attackers to execute arbitrary code via a malformed XML tag.

Stack-based buffer overflow in the image tooltip implementation in Trillian before 3.1.12.0 allows remote attackers to execute arbitrary code via a long image filename, related to "AIM IMG Tag Parsing."

Stack-based buffer overflow in Cerulean Studios Trillian before 3.1.10.0 allows remote attackers to execute arbitrary code via unspecified attributes in the X-MMS-IM-FORMAT header in an MSN message.

Buffer overflow in the Display Names message feature in Cerulean Studios Trillian Basic and Pro 3.1.9.0 allows remote attackers to cause a denial of service (crash) or execute arbitrary code via a long nickname in an MSN protocol message.

The AOL Instant Messenger (AIM) protocol handler in Cerulean Studios Trillian allows remote attackers to create files with arbitrary contents via certain aim: URIs, as demonstrated by a URI that begins with the "aim: &c:\" substring and contains a full pathname in the ini field. NOTE: this can be leveraged for code execution by writing to a Startup folder.

Buffer overflow in the AOL Instant Messenger (AIM) protocol handler in AIM.DLL in Cerulean Studios Trillian allows remote attackers to execute arbitrary code via a malformed aim: URI, as demonstrated by a long URI beginning with the aim:///#1111111/ substring.

Cerulean Studios Trillian Pro before 3.1.5.1 allows remote attackers to obtain potentially sensitive information via long CTCP PING messages that contain UTF-8 characters, which generates a malformed response that is not truncated by a newline, which can cause portions of a server message to be sent to the attacker.