The European Union might want it to be easier for police to obtain data, but that doesn't mean it'll be easy for officers to read that data. The European Parliament has proposed amended regulation that would not only require end-to-end encryption when available, but forbid backdoors that offer guaranteed access to law enforcement. EU residents need to know that the "confidentiality and safety" of their data is "guaranteed," according to the draft, and backdoors risk "weakening" that privacy. The proposal has to be approved by Parliament and then reviewed by the EU Council, so there's still a chance that the rules will be softened if and when the amendments pass. If they do clear, though, they could set up a conflict between the EU and countries that aren't so fond of encryption.
A new custom developed backdoor program has been used in highly targeted attacks against organizations from Taiwan, Japan, South Korea and the U.S. over the past year. Malware researchers from Symantec first came across the program, which they've named Dripion, in August 2015. However, due to its custom nature and sparse use, it has managed to fly under the radar since as early as November 2013. When their analysis began, the Symantec researchers believed Dripion was a local threat used against organizations in Taiwan, where most of its victims were found. However, since then, they have found computers infected with the backdoor in other countries as well.
In this paper we extend the classical notion of strong and weak backdoor sets for SAT and CSP by allowing that different instantiations of the backdoor variables result in instances that belong to different base classes; the union of the base classes forms a heterogeneous base class. Backdoor sets to heterogeneous base classes can be much smaller than backdoor sets to homogeneous ones, hence they are much more desirable but possibly harder to find. We draw a detailed complexity landscape for the problem of detecting strong and weak backdoor sets into heterogeneous base classes for SAT and CSP.
This work provides the community with a timely comprehensive review of backdoor attacks and countermeasures on deep learning. According to the attacker's capability and affected stage of the machine learning pipeline, the attack surfaces are recognized to be wide and then formalized into six categorizations: code poisoning, outsourcing, pretrained, data collection, collaborative learning and post-deployment. Accordingly, attacks under each categorization are combed. The countermeasures are categorized into four general classes: blind backdoor removal, offline backdoor inspection, online backdoor inspection, and post backdoor removal. Accordingly, we review countermeasures, and compare and analyze their advantages and disadvantages.
A new backdoor which affects the Apple Mac operating system has been discovered by researchers which claim there is a link to the OceanLotus threat group. According to researchers from Trend Micro, the MacOS backdoor is a persistent, encrypted sample of malware used for surveillance and data collection. The backdoor was discovered in a malicious Microsoft Windows Word document which has likely been distributed via email, potentially through spear and phishing campaigns. The document has been crafted to appear from HDMC, a Vietnamese organization which was established to promote national independence and democracy. If the document is opened by a potential victim, the user is asked to enable macros, which triggers the backdoor.