200 likes | 222 Views
Utilize trust economics tools to aid CISOs in making informed IT investment decisions based on technology, business priorities, and employee reactions. Enhance decision-making through policy knowledge bases and modeling tools.
E N D
semi-automated modelling of XACML policies Simon ParkinAad van Moorsel Newcastle University
setting Chief Information Security Officer (CISO) must continuously make IT investment decision should take into account: • technology • business priorities and impact • employee reaction and impact tools to help CISO trust economics (HP, ML) Aad van Moorsel, Newcastle University, DTI Trust Economics
trust economics tools Aad van Moorsel, Newcastle University, DTI Trust Economics
knowledge base and policies knowledge base: • existing base of policies, guidelines, e.g., ISO27k • add ontology for user behaviour, trade-offs, etc. • add calculators, models as appropriate • helps CISO to take into account the major elements when making decision interface to knowledge base through security policies: set of rules fed into knowledge base if I do X, what happens and what else should I do? Aad van Moorsel, Newcastle University, DTI Trust Economics
knowledge base and policies knowledge base (for integration) policies (to express decisions,interface with kb) Aad van Moorsel, Newcastle University, DTI Trust Economics
prototype: XACML Policy Editor Aad van Moorsel, Newcastle University, DTI Trust Economics
a prototype: XACML + Demos2K • IT Security officer uses XACML editor to write policies • tool feeds the policies in Demos2K (stochastic process algebra specification, solved using discrete-event simulation) • tool returns outcomes: loss of data, confidentiality, productivity, ... XACML • eXtensible Access Control Markup Language • OASIS standard • policy language + specified interpretation Aad van Moorsel, Newcastle University, DTI Trust Economics
a prototype: XACML + Demos2K demos2k XACML Aad van Moorsel, Newcastle University, DTI Trust Economics
a prototype: XACML + Demos2K demos2k Aad van Moorsel, Newcastle University, DTI Trust Economics
USB stick modelling Aad van Moorsel, Newcastle University, DTI Trust Economics
day in the life of a USB stick Aad van Moorsel, Newcastle University, DTI Trust Economics
USB roles Aad van Moorsel, Newcastle University, DTI Trust Economics
USB locations Aad van Moorsel, Newcastle University, DTI Trust Economics
example demos2k // Here both the FRIEND and TRAITOR cases are taken as being very similar - after all the // TRAITOR player is supposed to be indistinguishable to the FRIEND - and thus the // accounting should treat them broadly the same way. etry [who == player__FRIEND || who == player__TRAITOR] then { try [ binom(1, prob_reading_unencrypted_item) == 1 ] then { // unencrypted case successful_reads := successful_reads + 1; successful_transfers := successful_transfers + 1; // accidental archive of accessable material - i.e. USB_unencrypted_items try [binom(1, probAccidentalArchive) == 1] then { syncV(USBreveal, [USB_unencrypted_items], []); } etry [] then { hold(0); } // if player is actually a TRAITOR try [who == player__TRAITOR] then { Aad van Moorsel, Newcastle University, DTI Trust Economics
goals Aad van Moorsel, Newcastle University, DTI Trust Economics
experiments Aad van Moorsel, Newcastle University, DTI Trust Economics
evaluation XACML + Demos2K • prototype works (XACML as input to a stochastic model) to do: • formalise this • generalise this • is a policy language the right interface? • CISOs don’t normally use it • flexible enough as interface to knowledge base Aad van Moorsel, Newcastle University, DTI Trust Economics