+ All Categories
Home > Documents > IFIP SummerSchool2007 boldt

IFIP SummerSchool2007 boldt

Date post: 12-Sep-2021
Category:
Upload: others
View: 4 times
Download: 0 times
Share this document with a friend
18
Privacy-Invasive Software and Preventive Mechanisms Martin Boldt School of Engineering Blekinge Institute of Technology S-372 25 Ronneby Sweden [email protected]
Transcript
Page 1: IFIP SummerSchool2007 boldt

Privacy-Invasive Softwareand Preventive Mechanisms

Martin Boldt

School of Engineering

Blekinge Institute of Technology

S-372 25 Ronneby

Sweden

[email protected]

Page 2: IFIP SummerSchool2007 boldt

Blekinge Institute of Technology

SE-372 25 Ronneby

+46 455 38 50 00

www.bth.se/eng

The Spyware Problem (i)

According to Earthlink (an American ISP) 55% ofall Internet connected computers are infectedwith various kinds of spyware

Spyware exists because information has value

Spyware is a fuzzy concept without any properdefinition

The fundamental problem is the lack of standardmechanisms for managing users’ informedconsent during software installation

Page 3: IFIP SummerSchool2007 boldt

Blekinge Institute of Technology

SE-372 25 Ronneby

+46 455 38 50 00

www.bth.se/eng

The Spyware Problem (ii)

We use the Gatorsoftware as anexample of what usersface on the Internet

During the installationusers face an End UserLicense Agreement(EULA)

It contains 6,645words and is presentedin a small window

Would you read it?

Page 4: IFIP SummerSchool2007 boldt

Blekinge Institute of Technology

SE-372 25 Ronneby

+46 455 38 50 00

www.bth.se/eng

The Spyware Problem (iii)

The EULA reveals that the following programs areinstalled:

eWallet

Precision Time

Date Manager

Offer Companion

Weatherscope

SearchScout Toolbar

Such programs create large revenues for the theirdevelopers

Spyware corporations report annual revenues inexcess of $50 Million each

Page 5: IFIP SummerSchool2007 boldt

Blekinge Institute of Technology

SE-372 25 Ronneby

+46 455 38 50 00

www.bth.se/eng

Agenda

Introduction

The Spyware Problem

Privacy-Invasive Software

Preventive Mechanisms

Future Work

Page 6: IFIP SummerSchool2007 boldt

Blekinge Institute of Technology

SE-372 25 Ronneby

+46 455 38 50 00

www.bth.se/eng

Violates existing

laws

Software Behaviour

Mightviolate existing

laws

Confusingexperience

Followsrequiredpracticesand laws

Followsoptimal

bestpractices

Malicious Deceptive Questionable Acceptable Exemplary

Page 7: IFIP SummerSchool2007 boldt

Blekinge Institute of Technology

SE-372 25 Ronneby

+46 455 38 50 00

www.bth.se/eng

Violates existing

laws

Software Behaviour

Malicious Deceptive Questionable Acceptable Exemplary

Mightviolate existing

laws

Confusingexperience

Followsrequiredpracticesand laws

Followsoptimal

bestpractices

Spyware

Page 8: IFIP SummerSchool2007 boldt

Blekinge Institute of Technology

SE-372 25 Ronneby

+46 455 38 50 00

www.bth.se/eng

Privacy-Invasive Software (i)

We introduce the term Privacy-Invasive Software(PIS) instead of using the term spyware

Software that cause negative privacy implications

More descriptive and less negatively emotive asother terms such as spyware, evilware, badware,or hijackware

Even if we use the term “invasive”

We believe an invasion of privacy could be bothtolerable and requested by the users if fullytransparent

Page 9: IFIP SummerSchool2007 boldt

Blekinge Institute of Technology

SE-372 25 Ronneby

+46 455 38 50 00

www.bth.se/eng

Privacy-Invasive Software (ii)

Legitimate software is shown in white colour

Spyware in light grey

Malicious software in dark grey

ParasitesTrojansCovertsoftware

Lowconsent

Semi-parasites

Unsolicitedsoftware

Semi-transparentsoftware

Mediumconsent

Doubleagents

Adversesoftware

Legitimatesoftware

Highconsent

Severenegativeconsequences

Moderatenegativeconsequences

Tolerablenegativeconsequences

Page 10: IFIP SummerSchool2007 boldt

Blekinge Institute of Technology

SE-372 25 Ronneby

+46 455 38 50 00

www.bth.se/eng

Preventive Mechanisms (i)

Today’s anti-spyware tools use the same methodsto target spyware as anti-malware tools use tocombat viruses and worms

These methods are not optimal in the spywarecontext

“Innocent” software is negatively affected by anti-spyware tools

Page 11: IFIP SummerSchool2007 boldt

Blekinge Institute of Technology

SE-372 25 Ronneby

+46 455 38 50 00

www.bth.se/eng

Preventive Mechanisms (ii)

We believe that new and more user-orientedcountermeasures are needed

Mechanisms that inform users about how thesoftware affect them and their computer system

We put forward the idea of using collaborativereputation systems to inform users

Reputation systems are successfully used by forinstance Amazon.com, eBay.com and IMDb.com

Page 12: IFIP SummerSchool2007 boldt

Blekinge Institute of Technology

SE-372 25 Ronneby

+46 455 38 50 00

www.bth.se/eng

Preventive Mechanisms (iii)

Gather previous users’ knowledge about softwareand present it to the new user

Software reputations highlighted by such a systemis the same as users today gain from computermagazines and Web sites

We have developed a prototype tool, which isintegrated into Windows XP’s installation process

http://www.softwareputation.com

Page 13: IFIP SummerSchool2007 boldt

Blekinge Institute of Technology

SE-372 25 Ronneby

+46 455 38 50 00

www.bth.se/eng

Screenshot I

Page 14: IFIP SummerSchool2007 boldt

Blekinge Institute of Technology

SE-372 25 Ronneby

+46 455 38 50 00

www.bth.se/eng

Screenshot II

Page 15: IFIP SummerSchool2007 boldt

Blekinge Institute of Technology

SE-372 25 Ronneby

+46 455 38 50 00

www.bth.se/eng

Antagonistic intentions (i)

In an initial attempt to address antagonisticintentions from community users, we assigneach user a trust factor

A user’s vote impact depends on this value

The system also make use of a metareputation system that allow communityusers to rate each others’ contributions

Page 16: IFIP SummerSchool2007 boldt

Blekinge Institute of Technology

SE-372 25 Ronneby

+46 455 38 50 00

www.bth.se/eng

Antagonistic intentions (ii)

It would also be possible to allow users tosubscribe to the contributions from only apredefined subset of all community users,e.g. only a trusted subgroup

The prototype rely on simple e-mailaddresses for distinguishing betweendifferent users during registration

Other measures than e-mail addresses areneeded for identification

Page 17: IFIP SummerSchool2007 boldt

Blekinge Institute of Technology

SE-372 25 Ronneby

+46 455 38 50 00

www.bth.se/eng

System impact

The widespread use of such a system would helpusers prevent undesired software to covertlyinstall on their computer

The PIS classification is transformed in thefollowing way as the user is presented withinformation about software

ParasitesTrojansCovertSoftware

LowConsent

DoubleAgents

AdverseSoftware

LegitimateSoftware

HighConsent

SevereNegativeConsequences

ModerateNegativeConsequences

NegligibleNegativeConsequences

Page 18: IFIP SummerSchool2007 boldt

Blekinge Institute of Technology

SE-372 25 Ronneby

+46 455 38 50 00

www.bth.se/eng

The End

Questions?

More information is available at:http://psi.bth.se/mbo/


Recommended