ISF congress post 9: Extending security intelligence with big data

Extending security intelligence with big data

Presentation by Martin Borrett from IBM.

  • Why cyber security as a big data problem
  • How the diverse and rapidly changing set of both structured and unstructured data can play a key role in identifying the increasingly sophisticated threats that organisations face.
    • Move from reactive to a more proactive stance by actively searching for indicators that something could be amiss.

 

As an example, the attacks earlier this year on the New York times when it ran a story about China’s prime minister;

  • Not detected for 4 months
  • 45 different pieces of malware were used, with only 1 being picked up by AV
  • All employee passwords stolen
  • Computers of 53 employees accessed
  • University computers were used as proxies to hide the traffic source.

We have a greater need for security intelligence;

  • User identities
  • Assest discovery
  • Network flow
  • Vulnerabilities / risks
  • Security and threat feeds
  • Baselines of behaviour (system and user)
  • Unstructured data such as free text user inputs, feeds from social media, general news sources etc.

 

Attackers continuously adapting to leave minimal trace and hide their behaviour in the noise of ‘normal’ activity.  Due to the potential huge volumes of data, these systems must be very scalable.

Traditionally SIEM type solutions have focussed on real time alerting that is Proactive, Formalised (standard queries / searches) and fast.  This is great, but can it be in depth enough, and is real time attesting always required when searching for long term PAT style attacks?

Move towards adding more Asymmetric / Forensic type capabilities that are more Predictive, Inquisitive, and in depth.  These require considerably more skill and in depth understanding to create, and the searches will be much more ‘custom’, but this is the best (only?) way to find the subtle and clever attackers, especially if doing so in a timely manner is required (it is!).

Current SIEM type security processes may look like;

Screen Shot 2013-11-05 at 09.39.43

This has a heavy focus on structured data and performing real time correlation to get to a potential incident to investigate.

Moving more into the ‘big data’ world we will enrich this with a lot more data sources, much of it unstructured;

Screen Shot 2013-11-05 at 09.41.59

This will potentially also take outputs from the traditional SIEM tool as one of the feeds and enrich them with other data. An example may be where something that may be an issue, but where there isn’t enough detail to act on in the SIEM, this could be added to the ‘big data’ solution and correlated with a much wider data set to find out if it could be a real issue.

The top part of the above diagram (Real-time Processing and Security Operations) is relatively similar to existing SIEM solutions, focusing on real time analysis and processing, just with a potentially larger data set.

The bottom right (Big Data Warehouse, Big Data Analytics and Forensics) focuses on the much more advanced, not real time analysis and forensic type investigations.

Context is key.

  • You must be able to derive security relevant semantics from elements of the raw data.
  • There must be the capability to distil the huge volumes of data down to useful and real insights.
  • Human knowledge must ba able to be added to the solution to improve processing and automate more tasks.

Some key security questions a big data analysis solution will help your organisation answer include;

Screen Shot 2013-11-05 at 09.55.12

Another key area these tools can help with is in creating visualisations of attacks and suspicious behaviour.  As they will have data from all the systems in the enterprise, along with various external feeds, they can provide visual representations of the behaviour as it moves into that through the organisation.

For me the key consideration is to have one ‘Big Data’ solution that collects all the relevant data for your organisation from traditional log files, through corporate emails to social media and threat feeds.

This also needs to move out of the security realm as people are talking ‘Big Data’ but in reality still have the traditional SIEM mindset.  Running a tool like this for security, while the ops guys are also running logging and monitoring tools is massively wasteful in terms of cost, storage, management overhead, and also likely results in situations where some useful information only ends up in one tool, not both.

We need to move forwards to the mindset of an Enterprise ‘Big Data’ solution for sorting and correlating All the business data – logs, emails, external sources, user and system behaviours etc.etc.  This solution then has different dashboards, reporting solutions, search headers or whatever for the different use cases such as ops, business users (system performance, investigating transaction issues etc.) and ops.  Obviously areas like separation of duties and access controls must be considered here, but I believe this type of solution is the only way for this to really succeed and provide the best value for the business.

K

RSA’s First UK Data Security Summit – part 3: Defend with confidence against advanced threats

This talk covered three agenda items, with an obvious focus on RSA Security Analytics.

1. Why / how security investments need to shift

2. Building a SoC

3. Demo of the tool

Obviously I wont be capturing the Demo here, but below are my notes from this presentation;

Advanced threats are different…  Often following a similar set of steps;

– System intrusion – Attack begins – Cover-up discovery leap frog attacks – cover up complete, with the following characteristics;

  • Targeted
  • Stealthy
  • Interactive

How to defend;

  • decrease dwell time (time from successful breach until discovery)
  • speed response time (speed with which attacks are detected, and then remediated once discovered)

Relatively new attack discovered / named last year – ‘Waterholing’ – sit by the waterhole knowing prey will come to them – malicious users take over a site, knowing their targets are likely to visit it and trust it – then wait for them to arrive – malware etc. then delivered to users of the site.

Massive % of security spend currently on prevention, not detection..

71% of organisations have some sort of SoC (wider survey 66%)  most have plans to have one.  The question did cover from just some analysts who do investigations right through full on SoC capabilities.

SoC – level 1 adds, moves and changes, device health etc.

CIRC – manage security incidents, investigate suspicious behaviours, vulnerability analysis, threat management etc.

CIRC – even the specialists need to specialise!!

CIRCs can / should comprise the below 4 areas of responsibility.  Note, a person can have multiple roles, doesn’t need to be 4 people or more for smaller organisation1 – 4 suggested Tiers / areas of responsibility

  1. Front line – initial investigations, containment, triage, 24*7 etc
  2. Advanced tools, tactics and analysis – reverse engineering, host and network forensics, Cause and origin determination
  3. Analysis and tools support – Optimising the CIRC tools and processes; Integration, Content development, Reporting, Alert and Rule creation
  4. Cyber Threat Intelligence – understand the wider environment, analyse threat feeds, awareness of criminal / activist organisations etc.

EMC example – 1046 employees received a clear phishing email about fake wire transfers, 17 clicked on the link, 2 even clicked on the are you sure warning from the EMC gateway!  This sort of investigation should take minutes..  Does it for your organisation?

The maturity Journey – Control – Compliance – IT Risk – Business Risk

  • Your business needs to be moving from at least compliance to IT risk for levels 3 and 4 of the SoC to make sense.
  • Business, then IT risk SHOULD drive your security program and strategy.  Compliance is a byproduct of good security.
  • MSSP (Managed Security Service Provider)  – Make CIRC function more complete and affordable
    • What does it make sense to outsource from the CIRC functions?
      • Start with Tier 1, second most likely threat intelligence (as this can be somewhat stand alone, and an MSSP likely already has good contacts and threat intelligence they can share)
      • Tiers 3 and 4 can be, but these are harder and likely require in depth expertise and knowledge about the internal operation of the organisation.

To assist this organisations need;

  • Comprehensive visibility
    • view, collect and analyse everything
  • Agile analytics
    • efficient analysis and instigation of potential issues
  • Actionable intelligence
    • understand ‘normal’ aid identification and investigation of anomalies.  Make data machine readable
  • Optimised incident management

RSA Security Analytics is designed to meet these needs.  Well there had to be some product focus as it’s an RSA presentation..

My questions;

  • However, where does this fit into the overall business?
    • Can it be used by the wider business in order to offer a business wide solution to log management and analytics?

RSA response – Data is stored in Hadoop style storage so you can write tools to query it. But no there are no plans for them to provide any ops style dashboards and functionality that could be used by the wider IT team and the business.  For me this is a massive gap given the current market for log correlation and analysis type tools.  There is no way a business should want two of these solutions in place with logs shipped to both and all the associated licensing and management that goes with it.  Having two tools also leads to a potential situation where all logs may not get to the security tool and therefore you’ll miss potential threats.

So back to the talk;

RSA Security Analytics provides both a combination of both real time and longer term analytical abilities;

  • real time example – analysing data on the wire for attacks and suspicious behaviour
  • longer term – log on from two different locations – analyse distance between locations and time between logons 

Threat intelligence from feeds and incorporating business context. 

  • Look at all the data, use intelligence to narrow it down to provide a low number of real and useful alerts.

Security analytics demo;

  • Has full data set, can drill down to specific IP addresses, and the behaviour between it and others, identifies hacker tools etc.
  • Integrates with RSA threat feed etc.
  • Identifies high risk file types, windows cli commands etc.
  • Keeps suspicious IP address list from top suspicious IP list.
  • Can make network data back into the real data – e.g. can view emails as the email with cc etc, can view text files and images this looks a bit like man in the middle stuff – recompiles the actual conversation / traffic.
  • Currently a detective / investigative system.

5 take aways things you could do;

  1. Analyse current / goal security spend by prevention, detection and response.
  2. Honestly assess your organisations security maturity.
  3. Expand / build-out SoC/CIRC via on-premise or MSSP (or on premise MSSP).
  4. Invest in breach readiness processes.
  5. Evaluate your security tooling – is it too perimeter / signature based? Does it align with your security strategy and desired posture?

Overall this was a useful talk with quite a few good points and outside of the demo relatively little product and marketing talk.

I am however very disappointed that RSA are intent on keeping Security Analytics 100% focussed on security only.  It’s undoubtedly a good product in this space, but there are other products now that appear to offer similar levels of functionality in this space while also being genuinely good products across ops / application support / business users etc. and also being potentially more flexible and extensible.  Take a look at both Splunk and LogRythm.

K

RSA’s First UK Data Security Summit – part 1

On Monday I attended RSA’s first UK Data Security Summit at the Barbican.  Unsurprisingly this event had two main focuses;

– ‘Big Data’ – What it is, what it means to businesses and security, and how security can leverage it to look for anomalies and advanced threats.

– Security analytics – The relatively new RSA log correlation and analysis product.

The agenda from RSA was listed as;

  • Big data and the hype
  • The changing threat landscape
    • Cyber criminals, nation states, activists and terrorists
  • Balancing risk of attack and prevention against ability to perform key tasks

As with my recent Splunk Live! post, the below will be relatively unformatted, but hopefully still of use.

The day started with some keynote talks from Art Coviello, Eddie Schwartz and Andrew Rose;

Art Coviello – Intelligence driven security: A new model using big data

Arts’ talk focused on the rapid changes to the IT environment over the last few years, with predictions for the future as well, then moved into the historic and current security  model and what this needs to look like in the future.

70’s – terminals – 1000s users

90’s – PCs – millions users

2010 – Mobile Devices – billions users

Digital content;

2007 – 1/4 Zettabyte

2013 – 2 Zettabytes

2020 – 100 Zettabytes

5* more unstructured than structured data, and growing 3* faster.

Apps;

2007 – web front end apps

2013 – Theres an app for that

2020 – big data apps everywhere..

Devices;

2007 – Smart phones

2013 – dawn of really smart phones and smart phone / tablet ubiquity

2020 – Internet of things (everything from fridges to coke machines as well as all the usual phone / pc / tablet etc devices)

Social media

2007 – MySpace

2013 – Focus on monetizing

2020 – Total consumerisation of social media: absence of privacy..

Perimeter;

2007 – holes

2013 – is there a perimeter?

2020 – no direct control over physical infrastructure..

Threats;

2007 – Complex intrusion attacks

2013 – Disruptive attacks – can’t launch physical attacks over internet yet, but can be very disruptive

2020 – Destructive attacks? with no physical / user interaction required?

Historic security model;

  • Reactive
    • Perimeter based
    • Static / signature based
    • Siloed
      • Firewall, IDS, AV etc – all reactive, don’t play together or support each other

New model;

  • Intelligence driven
    • Risk based
    • Dynamic / agile
    • leveragable / contextual
      • Look for anomolys, be more heuristic / intelligent, work together – correlate events across the enterprise

Impediments to change;

  • Budget inertia: reactive model
    • 70% on prevention (likely more like 80 % in many firms)
    • 20% Detection and monitoring
    • 10% Response
    • Skilled Personel shortage
    • Information sharing at scale – industry groups, sharing data of attacks and breaches etc at ‘wire speed’
    • Technology maturity
      • Some commentary about archer, silver tail etc. RSA has bought or invested in

Look at security maturity model;

  • Stage 1 – Unaware (wish security would go away, install a box to fix it all)
  • Stage 2 – Fragmented (compliance gathering – focus on box ticking to get compliance rather than doing security right)
  • Stage 3 – Top Down (security understood but driven from management down, not yet pervasive)
  • Stage 4 – Pervasive (good security team, work with c-level on budgets etc)
  • Stage 5 – Networked  (working across the business and integrated with the business)

Big data transforms security;

  • Security management
    • Scalable to analyse all data
    • generates a mosaic of information
    • accelerates responsiveness
  • Controls
    • task specific
    • behaviour orientated
    • self learning
  • enables view of attacks in real time

Need this detailed analysis in order to prevent / see sophisticated attacks such as man in the middle and man in the browser

Intelligence driven security needs to be resilient, feed into controls and in and out of GRC stack (grc feeds into and educates controls.  controls feed into GRC to confirm compliance)

 

Eddie Schwartz – Embracing the uncertainty of advanced attacks with big data

Pecota forcasts – analytics platform used by bookies to work out odds one sports / sports players – baseball – movie – money ball.

– ‘big data analytics’ changed the way baseball players were assessed and consequently paid..

Facebook data mines images as well as text on your page to drive targeted advertising

Amazon etc. – preference engine – you bought this, you want these..

* They are information rich and using high quality analytics.  Why are we not using data like this in security?

Why? – too much time having to say yes we are ok, yes we pass xx audit..

Attackers do not have these checklists – they will work hard to breach any opening regardless of whether you are complaint with whatever regulation..

  • Read ‘the signal and the noise‘ – Nate Silver – why so many predictions fail and some don’t.
    • The signal is truth, the noise is what distracts us from the truth.

How much do we really know about our adversaries?

  • Are we researching the tools, techniques and processes of our adversaries
  • Do we know who they are?
  • Insiders, hackers, hactivists, criminal organisations, nation states etc.
  • Do we know what they look like?
    • Old world (SIEM) – finite, rule sets, wait for rule to be breached
    • New world – infinite – unknown unknowns, uncertainty, hackers may look like legitimate users – what signs can we look for to identify them?
  • Do we understand the ‘Kill Chain’ – Prepare, Infect, Interact, Exploit
    • Cost to remediate goes up dramatically as you move along the chain
    • detection sweet spot – when they first exploit / attempt to exploit – they have to reveal themselves, so fast detection here will catch / print before data exfilitration.

Need to move to more spend and more intelligence on ‘internal’ protection / detection / capture – away from the traditional perimeter.

What are your drivers for IT security investment?

34% compliance, 16% audit

ONLY 6% strategy!

Big data transforms security – 4 areas for shift..

  1. Security management
  • Comprehensive visibility – not just event logs – what are my critical processes, what information do I need to see to understand if they are at risk.
  • Actionable intelligence – must be available in a timely manner
  • Agile analytcs – security environment must be able to change as the environment changes – your environment is at least somewhat unique, also threat landscape changes
  • Centralised incident management – can security teams follow an incident from end to end? – many point solutions.. Do logs all go to one place, can they be effectively analysed?

2. Intelligence driven security

    • Ah-hoc – Bystander – End User – Creator; Crawl – Walk – Run – Advanced – World Class
    • Monitoring and detection, incident response, threat intelligence, systems and analytics; Where should we be – risk based – do you need to be world class in everything? Where do we need to focus, what are our risks?
    • Critical Incident Response Centre (CIRC) – Cyber threat intelligence, Advanced tools, tactics and analysis; Critical Incident response team, Advanced specialists

3. Live intelligence

 

  • Threat intelligence, rules, parsers, alerts, feeds, apps, directory services, reports and custom action.
  • Need long term technology, process and architecture plans
  • Visibility, control, governance, intelligence are all interrelated and must be considered as parts of a whole.

4. Risk based authentication

 

  • Active input – username, password, one time password, certificate, out of band, security questions, biometrics
  • access time, access location, geo location by IP, location by access point,
  • What does ‘good behaviours’ look like vas. ‘bad behaviour’; profile behaviour
  • Criminals cannot replicate your unique use profile.
    • Velocity, page sequence, origin, contextual information; velocity, behaviour, parameter injection, man in the middle, man in the browser.

Shift discussion in GRC from meeting compliance regulations to focusing IT and security staff on the key work

  • right assets and processes based on criticality and importance
  • assest intelligence, threat intelligence, event focus, investigations – Analyst prioritisation
    • requires accurate, timely and complete data.
  • read – Big data fuels intelligence driven security – RSA white paper

US – Data sharing bill – both businesses and liberal groups have objected.

  • how to share without compromising privacy.
    • criminals already violating our privacy every day
    • who should protect our privacy – benign government, corporations, criminals?
    • laws protecting customer privacy can make it hard not to breach laws protecting employee privacy in the EU?

 

Andrew Rose – principle analyst – security and risk management – Forester – ‘An external perspective’

Information classification – how mature

  • 26% have a policy that’s widely ignored, 28% have a policy for some data or systems..

The world we live in (largely as previous presentations)

  • Increasingly capable attackers (threat is real – activists, china etc..)
  • Budgets relatively static or slow growth, enough for triage of known issues, not whole treatment and improving security posture.
  • ROI – hard to define / prove – if not breached are we good or just lucky.  No good model seems to exist yet.
  • Yes rather than no security culture – have to work with business and enable – increase risk and complexity to deal with, but not necessarily staff and budget..
  • Competitive recruitment environment
  • Even the best firms have flawed security – e.g. RSA breach – have to prepare to fail!

Forester and IBM reports has IT at the top of the list of most important reasons for business success.

However business and IT (business especially) do not rate the success / competency of IT very highly – not agile, can’t accommodate change, can’t deliver projects on time etc.

 

RSA yearly IT security challenges included;

  • Third highest issue (76%) – changing business priorities
  • Forth (74%) – day to day tasks taking too much time
  • 8th (55%) lack of visibility of security – fixing this one will likely improve other issues at lot.
  • adoption of ISO / cubit etc not helping these keep getting higher up the issues scale

 Business innovation does not slow down because of security threats…

Complexity vs. manual ability – can better analytics help?

Vendors – vendor space is buzzing..

  • security commercialisation is in full swing
    • But what are the differentiators – everyone users the same buzzwords to sell products (e.g. big data, threat intelligence etc.)
  • Disruptors needed
    • need innovation, not re-hash or updates
    • services, not more hardware
  • solutions fragmented
    • how many products required to ‘solve’ security
    • what do I need now
    • what order should I buy them
    • what is the value / roi?
    • how much resource does it take to manage?
    • too many niche products – e.g. IAM, remove admin rights etc.  Need a ‘BIG’ tool / solution, to solve many / most issues and integrate existing products / solutions.

SIEM

5% get great value, 30% have not implemented, 65% get little or limited value

So is Big data the solution?

  • Big data just means lots of high velocity, structured and unstructured data – it is there to be used – so it is what you do that counts with it, not it in its self (my comment, not speakers)
  • supply chain complexity
  • technical complexity
  • internet of things

 

For me same conclusion as before – need something to aggregate and bring all the data together from apps, security tools, systems and then analyse it.  intelligent, fast correlation – look for real connections and real relationships – be mindful of coincidences in the noise.

 

2 books – anti fragile, signal to noise.

Common pitfalls –

  • starting with the data – need context and understanding as well.
  • overlooking the value of metadata.  data tagging increases value of data
  • believing more data is better
    • think simplicity and actionability

 Take away points;

  • Understand and identify your data
    • information classification is key – get this accepted and rolled out across the business
  • Be ‘hypothesis-led’ – think of what you cold do, not just what you know – then see if you can find the data to achieve it
  • Look for business partners for any big data initiative – again – one engines / dwh etc.

I’ll complete my write up of the day shortly, I hope you’re finding it useful.

K