Spectre and Meltdown: What it really means
Posted by Peter Bassill on 08/01/2018Over the past few days, you will have heard of the new vulnerabilities affecting Intel and AMD processes, but what does it really mean?Named Spectre and Meltdown, there has been a lot of hype in the press about them "being the end of the computing environment” and “all processors should be replaced”.Over the past few days our CEO, Peter Bassill, has been looking into the issue and working on some proof of concept code to understand the exposure more fully.As the issue is gaining a lot of exposure and a lot of questions are being asked, here is an interim security bulletin, with a full report ready for the end of the week.Both vulnerabilities require a highly complex attack which is only locally exploitable. This is fine if the servers are your own, as an attacker would first need to compromise a system or network in order to gain access to the system. Once on the server then the risk is increased. For cloud provisioned servers the issue is greater because if it is possible to perform the exploit on a local account on the cloud server then it would affect all users.What do I need to do?For the moment the only action available is to patch systems. Where cloud provisioned servers are used, a risk led decision should be taken to whether the system should be brought internally. Certainly, where confidential information is held on a cloud provisioned server we would strongly recommend bringing this server onto organisationally owned infrastructure.Predictions for the next monthBoth of these vulnerabilities presently pose a low risk to organisations but that will significantly change as attackers have more time to develop exploit code.We fully expect to see exploit code and attack modules being made public in the coming weeks. Once the exploits become more available, the Real World Risk score will increase to a predicted 70/100 for cloud provisioned systems and 54/100 for private servers.