Send the link below via email or IMCopy
Present to your audienceStart remote presentation
- Invited audience members will follow you as you navigate and present
- People invited to a presentation do not need a Prezi account
- This link expires 10 minutes after you close the presentation
- A maximum of 30 users can follow your presentation
- Learn more about this feature in our knowledge base article
Do you really want to delete this prezi?
Neither you, nor the coeditors you shared it with will be able to recover it again.
Make your likes visible on Facebook?
Connect your Facebook account to Prezi and let your likes appear on your timeline.
You can change this under Settings & Account at any time.
BeeWise: A Futures Market for Fostering Security by Design
Transcript of BeeWise: A Futures Market for Fostering Security by Design
our society will live only the future it will choose to live into. Bugs & Carrots M&Ms Bees & Crowds Back To The Future "Security metrics are the servants of risk management, and risk management is about making decisions. Therefore, the only security metrics we are interested in are those that support decision making about risk for the purpose of managing that risk"
- Dan Geer, foreward to Andrew Jaquith's Security Metrics Today, security stakeholders face a challenging task in assessing the risks they are exposed to, as they have incomplete information about the number and the severity of vulnerabilities affecting their systems. BeeWise: "software products have among the highest levels of defects of any product sold today, and there is very little accountability on the part of producers and software products" Homeland Security Strategy for Critical Infrastructure Protection
in the Financial Services Sector, May 2004 Software makers Software Buyers Hackers Economic Security Metrics Two main areas of research in economic approaches for security metrics:
One has its roots in investment and decision theory and is mainly pursued in the field of information technology-oriented business administration. It has yielded a number of quantitative metrics that can be applied as guidelines in investment decisions as well as for the evaluation of existing security measures.
The second area of research has ancestors in micro-economics. It deals with market concepts to gather security-relevant information and extract quantitative indicators on information security properties. New Incentives when people engage in activities that impose high social costs, it usually means that private cost is too low...
... to change behavior then, requires raising the private cost of a particular activity. Internalize the Externalities Establishing laws and regulations in the information security area (i.e., fixing liability models, security will not be perceived as an externality anymore);
Establishing new markets with feedback mechanisms (i.e., balancing information between buyers and sellers, hence mitigating the problems at their source) Two Perspectives The ex-ante perspective tries to assess the costs and benefits of possible future investments and helps to decide whether an investment project is profitable or not. It helps to decide whether to invest in a certain security measure or not, or to choose the best alternative out of different possible security measures ("What measures should we implement?").
The ex-post perspective for the retrospective judgement of past investments. It provides a target-performance comparison and answer the question if the firm's resources were spent effectively ("Did we do the right things?"). Metrics for IT Security Investments A number of Investment Metrics has been suggested and adapted to the IT Security context. They include:
NPV (Net Present Value);
IRR (Internal Rate of Return);
VAR (Value at Risk) - adapted by Juninger et al.
Limitations exist and also challenges... Challenges in Applying Investment Metrics in IT Security Reliable data: risk management metrics require estimations for severity and probability of loss events. In traditional risk management those values are often deducted from historical data. This is barely feasible at the moment;
Quantification: Historical data is scarce even within organisations that have been using IT technology for a long time because it is hard to quantify data or for a lack of rigor of methodological collection;
Complexity: it is hard to really get a comprehensive picture of possible threats to an organisation's security. Metrics Based on Market Mechanisms Metrics based on market mechanisms are not necessarily employed in an economic or business context, but that their way of measurement is based on economic principles;
The observation of economic agents' decisions yields useful indicators for their expectations and can eventually be used to construct operable metrics;
Market prices are not always accurate for different reasons (eg., low liquidity, transaction volumes, false estimations);
They provide a good approximation;
These kind of metrics are forward-looking because market prices are based on expectation about the future rather than on historical data. Metrics Derived From Existing Markets Using standard approach in finance research called event study, a number of studies analyzed the impact of computer security (and privacy breaches) related news on financial market prices, typically the stock price of affected or involved publicly traded firms;
Assuming the absence of a publication bias, particularly those event studies with large samples provide evidence for a market impact of information security news;
There are limits to the adoption of stock market prices as a direct metric for security properties, namely the inability to capture medium and long-term losses (only short term ones get captured) and dependance on the occurrence of extreme events;
They are merely a Post-HOC indicator for insecurity, rather than a metric for security in a state where no incident happens. The development of stock market prices before the event of interest - the estimation window - is fed into a prediction model for each individual stock.
Then the prediction model is employed to forecast the most likely development of the stock price after the event date.
It thus simulates the scenario as if the event would not have happened.
The abnormal return, defined as difference between estimated and actual returns during the event window, can be interpreted as a measure for the impact of the event on the market valuation of a firm. "Due to several economic reasons, namely network externalities and information asymmetries, neither vendors have incentives to build sound security technology into their products nor are users willing to spend extra money on security technology"
Ross Anderson, Why Information Security is Hard Well intentioned actions can have unintended consequences Perverse
Incentives Software Maker
Cost Patch Cost to Buyer Patching Features and
Information Asymmetry Information Seller's
Knowledge - RESULT: A depth spiral
- High quality, secure software is
driven from the market
- Lower-quality, feature-rich software tends to dominate
- Less scrupulous parties enter the market (software pirates)
- Individual behavior makes everyone worse off Market Failure, at present SUPPLY: Lemons market suggests that vendors under-supply security to the market;
DEMAND: The tragedy of commons tells us that users demands less security than appropriate; What is Security Today, from an Economic Perspective How to invert it?
How do we change? Design of Specific Information Security Market Even though the actual extent of the security market failure and the relative contribution of technology and policy are difficult to gauge, many of the countermeasures proposed in the literature stimulate new markets and therefore are not only good tools to align incentives, but also to obtain a new class of security metrics. Vulnerability Markets: A Taxonomy The basic idea:
Vulnerabilities are errors in computer systems which can be exploited to breach security mechanisms. As a security-related information, vulnerabilities can be traded.
exploit Derivatives. How Vulnerability Markets Compare? An ideal vulnerability market fulfill three, plus one, functions:
Information function – the ability to use market prices as forward-looking indicators of security properties (i.e., countering the lemons effects);
Incentive function – allow monetary compensation for security research and development (high priority to security issues!);
Risk-balancing function – the market provides instruments to hedge against large information security risks (i.e., mitigate the impact of occasional security breaches);
Efficiency – orthogonal to the other functions and characterized in terms of: low transaction costs, liquidity, transparency, accountability. How Vulnerabilities Market Compare? With Bug Challenges and Auctions there is no possibility to do risk-balancing at all; moreover the information obtained about the market price is only a lower bound;
Brokers gives questionable incentives and not always disclose vulnerability (i.e., information) to the public;
Exploit Derivatives and cyber Insurances looks both promising. The former provide a timely indicator, while insurances can be less efficient due to high transaction costs, bad portfolio balancing/high correlation risk. Exploit Derivatives Transfer the mechanism of binary options from finance to Information security;
No sensitive vulnerability information gets traded;
The market is build around contracts that pays out a defined sum in case of security events; Exploit Derivatives: Market Participants A larger number of interest groups get attracted!
Security experts and vulnerability researchers. Exploit Derivatives: Contracts Consider a pair of contracts (C; C'), where C pays a fixed amount of money, say 100 EUR, if there exists a remote root exploit against some specied server software X on platform Y at date D in the future;
The inverse contract, C' pays out the same face value if there is no remote root exploit submitted to a market authority before date D;
It is evident that the value of the bundle (C; C') is 100 EUR at any time and that selling and buying it is risk-free Exploit Derivatives Assume that there is an exchange platform, where the contracts C and C' can be traded individually at prices determined by matching bid and ask orders. The platform settles the deals, and publish the price quotes from the order book;
Then the ratio of the market price of C and its face value approximately indicates the probability of software X being compromised before date D;
Of course liquidity, and thus a high number of market participants as well as low transactions costs, are a prerequisite for this mechanism to provide timely and accurate estimates. Exploit Derivatives: Software Users Software users would demand C in order to hedge the risks they are exposed to due to their computer systems in place. Exploit Derivatives: Cyber-Insurance Companies The same applies for cyber-insurance companies underwriting their customers' cyber-risks. Exploit Derivatives: Investors Investors would buy contracts C' to diversify their portfolios. Exploit Derivatives: Criticalities The Exploit Derivatives market requires a Market Authority to test candidate exploits at the end of each contract and announce the results;
The Market Authority can be asked to publish the exploits with the announcements, providing verifiability and countering fraud;
The Market Authority role can be distributed to different actors;
Modeling contracts (eg, the importance of reliability growth models, understanding defect density, distribution, and nature)
Weather derivatives were first traded by companies accustomed to trading contracts based on electricity and gas prices in order to hedge the prices risk of their utilities. How to bootstrap exploit derivatives? Exploit Derivatives: Software Vendors Software vendors could demand both types of contracts: contracts C that pay if their software remains secure as a means to signal to their customers that they trust their own system;
or contracts C_comp that pay if their competitors' software gets compromised.
One could even think of software vendors using exploit derivatives as part of their compensation schemes to give developers an incentive to secure programming. Exploit Derivatives: Security Experts could use the market to capitalize efforts in security analyses.
If, after a code review, they consider a software as secure, they could buy contracts C' at a higher rate than the market price.
Otherwise they buy contracts C and afterwards follow their preferred vulnerability disclosure strategy. Exploit Derivatives: Remarks No co-operations of the vendor is needed;
The number of different contracts referring to different pieces of software, versions, localisations, etc., is solely limited by demand;
They are expectation-based and forward-looking
since almost all markets are settled in money, the often-dicufficult quantication is inherently implied in the market mechanism;
On the downside, while most existing markets are good at predicting future states in the long run, the advantages go along with short-term frictions and, in the terms of metrics, with measurement error Weather Derivatives First apperead in 1996 and 1997 US energy industry
They allows businesses and other organizations in insure themselves against fluctuations in the weather
Hedging with weather derivatives reduces the year-to-year volatility of trading companies, Exploit Derivatives vs
Play Money Prediction Market Common Vulnerability Scoring System AV:[L,A,N]/AC:[H,M,L]/Au:[M,S,N]/C:[N,P,C]/I:[N,P,C]/A:[N,P,C]
CDP:[N,L,LM,MH,H,ND]/TD:[N,L,M,H,ND]/CR:[L,M,H,ND]/ IR:[L,M,H,ND]/AR:[L,M,H,ND] Base
Environmental BeeWise Symbols BEE Symbol:
WIN7-6.13-AV:N-RC:C Examples: Relation of events and price quotes of hypothetical exploit derivatives (security-event futures) What Does it Take... Heterogeneous user base;
Low transaction costs source: Rainer Bohme, A Comparison of Market Approaches to Software Vulnerability Disclosure DNET4x-6.13-AC:H-RC:C The contract pays $100 BEE dollars to the holder, if a vulnerability requiring specialized access conditions (AC:H) - as defined by CVSS - is found to affect the Microsoft .NET Framework 4.x by June 30th 2013 (6.13). This contract requires (RC:C) the vulnerability to have been acknowledged by the vendor, or confirmed from an external event such as the publication of functional or proof-of-concept exploit code or widespread exploitation. OBSD-6.13-Au:N-C:P,C-RC:C There are allegations about an FBI-planted back door in the OpenBSD IPSEC stack: http://marc.info/?l=openbsd-tech&m=129236621626462&w=2
The contract pays $100 BEE dollars to the holder, if the allegations will be confirmed (RC:C), by June 30th 2013 (6.13). The contract is related to vulnerabilities:
not requiring the attacker to authenticate in order to exploit them (Au:N), and
partial or complete impact on confidentiality (C:P,C), and
acknowledged by the vendor or software authors, or confirmed from an external event such as the publication of functional or proof-of-concept exploit code, or widespread exploitation Markets Selection Criteria Prediction Markets: Basic Idea virtual stocks on an electronic market whose pay-offs are tied to the outcome of uncertain future events;
Although the final pay-offs of stocks are unknown during the trading period, rational and risk neutral traders sell stocks if they consider the stocks to be overvalued and buy stocks if they consider the stocks to be undervalued;
As a result, the trading price reflects the traders’ aggregated beliefs about the likelihood of the future event. Market prices can thus be interpreted as predictions;
Hayek hypothesis, theoretical justification: asymmetrically dispersed information is best aggregated using a price mechanism. 200 John Brunner publishes a science fiction novel, Shockwave Rider, with a description of a prediction market that he called the Delphi Pool 1975 Prediction markets are introduced to a much wider audience by the popular book The Wisdom of Crowds by James Surowiecki 2004 How to measure Anything - Finding the Value of the Intangibles in Business - by Douglas W. Hubbard 2006 2001 - DARPA Information Awareness Office research the possibility of using prediction markets in policy analysis;
2002 - demonstration markets about the spread of SARS and security threat level;
2003 - Policy Analysis Market Program gets cancelled, after critics savaged DARPA for proposing "Terrorism Features" 2001-2003 DARPA "Terrorism Market" Affair Prediction Markets: Does Money Matter? Prediction accuracy: market forecast winning probability and actual winning probability with play-money and real-money - src: SERVAN-SCHREIBER et al. Prediction Markets: Does Money Matter? Prediction Markets: Does Money Matter? psychological importance of hard currency;
Geographical, financial, technical, fiduciary, regulatory, and, perhaps, ethical obstacles to the establishment of real-money predictions markets;
Does the play-money alternative compromise accurancy? Does Money Matter? Two Studies Prediction Markets: Does Money Matter?, Justin Wolfers, Emile Servan-Schreiber, David Pennock and Brian Galebach, Electronic Markets, 14(3), September 2004;
An Empirical Investigation of the Forecast Accuracy of Play-Money Prediction Markets and Professional Betting Markets, Slamka, C.; Luckner, S.; Seemann, T.; Schröder, J., European Conference on Information Systems (ECIS), 2008 Two Real-World Experiments real-world online experiment contrasted the predictions of TradeSports.com (real money) against those of NewsFutures.com (play money) regarding American Football outcomes during the 2003–2004 NFL season;
Empirical study that compares the forecast accuracy of a play-money prediction market for the FIFA World Cup 2006 to predictions derived from odds issued by two professional betting companies; Does Money Matter? Both types of markets exhibited significant predictive powers, and remarkable performance compared to individual experts;
Play-money markets performed as well as the real-money markets (neither type of market was systematically more accurate than the other);
play-money prediction markets found about as accurate as betting market;
real-money markets may better motivate information discovery (i.e., incentive function);
play-money markets may yield more efficient information aggregation (i.e., efficiency); Prediction markets starts to be available online to the general public, in both real-money (gambling) and play-money (game) mid 1990s Prediction markets outperformed forecasts by experts and opinion polls, in:
business related fields
current events 1990s - today Hayek hypothesis:
asymmetrically dispersed information is best aggregated using a price mechanism. 1945 Historical incarnations of prediction markets - Wagering on Presidential Elections
Although election betting was often illegal, the activity was openly conducted by “betting commissioners” and employed standardized contracts that promised a fixed dollar payment if the designated candidate won office 1868-1940 Prediction Markets The Road Ahead Experiments, Experiments, Experiments: including: non-negligible bid-ask spread, liquidity, pay-off functions, short-term frictions;
BeeWise traders competition;
Platform Development - a lot still to be done. Few topics: usability, mobile, ...;
Call for Participation: researchers, developers, investors. Interested? Please, be in touch. Information Seller's
Knowledge What will be the future for software security?
what will be its place over the next 10-20 years?
Will we find the way to balance the information between buyers and sellers in the security market?
Which incentives will we manage to provide to manufacturers to build security in? I believe the understanding of how humans behave, from an economic and psychological perspective, to be key to holistically address the security challenges and make the software the new foundation our society can rely on. Ship... ... then test Metric Value Description Weather Derivatives Instrumental in hedging against the exposure to certain risks;
They both differ from other derivatives wrt the underlying assets (i.e., information security events, weather events) which have no direct value to price the information-security/weather derivative; In turn, low volatility in profits can often reduce the interest rate at which companies borrow money;
in publicly traded companies, usually translates into low volatility of the share price, and less volatile shares are valued more highly;
low volatility in profits reduce the risk of bankruptcy. Combining information from more than one contract allows for even more interesting metrics.
Using Spreads: Price differences between related contracts can be directly attributed to distinctions in security (or public scrutiny) due to underlying technical
differences (Boheme). joint probabilities of failure can be computed from pairs of contracts to measure the total security of layered defense mechanisms.
How much survivability we get with defense-in-depth? Security-Event Futures: Remarks No co-operations of the vendor is needed;
The number of different contracts referring to different pieces of software, versions, localizations, etc., is solely limited by demand;
They are expectation-based and forward-looking;
since almost all markets are settled in money, the often-difficult quantification is inherently implied in the market mechanism. Two market types are supported:
Binary markets: where contracts pay the maximum settlement value to the holder, or nothing, depending on the occurrence of the underlying event by the contract closing date;
Index based markets: contracts pay a variable amount, depending on the pay-off function, taking the index as its input. Market Types Binary Markets A contract may be created that pays one dollar to the holder depending on whether the Linux kernel will be found to be vulnerable to a remote code execution by the end of the month.
Index-based Markets: A contract may be created that pays 1 BEE/USD cent for every data breach disclosed by a Fortune 1000 in the next year.
Index-based markets: A contract may be created that pays 1$ BEE/USD for every point the Index of Cybersecurity values above/below a strike K. Market Types: Examples This is why when we talk about the future of software security we should correct ourselves and say 'futures' (plural!).
This is why I believe security-event futures might contribute towards establishing information symmetry. Pay-offs Functions p(x) = min(L$, max(D(x - K), 0)) p(X) = min(L$, max(D(K - x), 0)) p(X) = max(-L$, min(D(x - K), L$)) p(X) = max(-L$, min(D(x - K1), max(0, min(D(x - K2), L$)))) p(X) = min(L$, max(D(K - x), min(D(x - K), L$))) p(X) = min(L$, max(D(K1 - x), max(0, min(D(x - L), L$)))) 'Patching allows software manufacturers
to optimize market and legal protections
by "re-negotiating contract terms buyers
could not negotiate in the first place' - David Rice 1st Part Inspired by David Rice Software manufacturers will not forgo markets share
Software buyers will not forgo features
Cyber attackers will not forgo attacking tens of millions of vulnerable systems Configuration Self Correction? Tragedy of Commons Both risks and benefits are socialized between all the elements of the population, individuals lack the incentive to unilaterally invest in security.
(see Garrett Hardin’s tragedy of commons) Information Seller's
Knowledge A Futures Market for Fostering Security by Design IFIPTM 2013, June 4th - NESSoS Industry Session