Cyber Threat Intelligence: Sharing Is Caring
Shared cyber-threat intelligence will soon be a critical component of security operations, enabling organizations to better protect their digital assets and respond more quickly to emerging threats.
Recently, the US Department of Homeland Security announced the deployment of the Automated Indicator Sharing (AIS) system, which allows the exchange of cyber-threat intelligence among private and public organizations. Increasing the breadth and speed of information sharing will reduce the number of security compromises, enabling all types of organizations to better defend themselves against emerging threats.
There is almost unanimous agreement among security professionals that cyber-threat information is valuable to their organizations. However, as we dig deeper into the attitudes and implementation barriers to sharing that information, we find myths and significant reticence.
First, let’s define cyber-threat intelligence and dispel a significant myth. Cyber-threat intelligence comprises details and metadata about suspicious and malicious activity, including attack vectors, weaknesses that are being exploited, and mitigation or containment actions. It does not contain any personally identifiable information, even when sharing a file reputation.
Next, let’s look at which threat and reputation data people are willing, and unwilling, to share. Intel Security recently surveyed almost 500 security professionals globally and found that about three-quarters of those involved with and knowledgeable about cyber-threat intelligence sharing are willing to pass on information about the behavior of observed malware. Malware details have been shared for a long time, typically with an incumbent vendor or nonaligned security organization. What is surprising is that this figure is not closer to 100%.
Around half of the security professionals surveyed are also willing to share reputation info on URLs, external IP addresses, and security certificates. This increased reluctance to share is typically attributed to company policy or industry regulations and often comes from concerns about legal repercussions from the entities that are identified as being potentially malicious.
Finally, only about one-third are willing to share file reputations, probably due to concerns about accidentally releasing some sensitive or confidential information in the file. Yet cyber-threat intelligence-sharing systems calculate a unique one-way hash to represent the file that is being convicted -- this is the only data that leaves the corporate system -- and the file cannot be recreated in any way using this value.
Sharing More Valuable Than Secrecy
Increasing support for cyber-threat-intelligence technical standards will help people understand exactly what is and is not included in a threat record and will broaden industry implementations. Although some organizations believe they stand a better chance of identifying and catching bad guys by themselves if they keep the attack details private, more and more realize that the changing nature of attacks makes sharing more valuable than secrecy. Standardization will also make it easier to combine and correlate multiple discrete observations into a larger and more accurate picture of a particular threat.
Catching modern, adaptive attacks is difficult for traditional endpoint and firewall defenses working in isolation because the attacks often mutate every few hours or days, faster than signature updates and scanning tools can keep up. The trend toward targeted attacks is also increasing interest in industry-specific cyber-threat intelligence. Although there are still barriers to overcome before cyber-threat intelligence sharing is widespread, those barriers are falling as successes are publicized and regulations are enacted to provide liability protection. Within a couple of years, shared cyber-threat intelligence will be a critical component of security operations, enabling organizations to better protect their digital assets and respond more quickly to emerging threats.
DarkReading: http://ubm.io/1ZR2gXZ