Analytics and Machine Learning Can Help Detect Cyber Threats

Wednesday, March 22, 2017

The challenge of cyber security is laid bare when RSA, one of the world’s preeminent security firms providing solutions to address cyber threats, is unable to prevent a social engineering attack on itself. This of course happened in 2011, but given the frequency with which companies are breached today, no enterprise can be a 100% sure of preventing an attack. Why? The real threat lies ‘within’ the organisation. Employees are the weak link and hence the easy target for cyber criminals to exploit in order to get hold of sensitive data.

So, how must a law firm defend against its own employees – the very people who it must provide data access to – without ‘getting in the way’ of their work? Furthermore, it’s not always apparent that the organisation is hacked. The average time that an attack went undetected in a network in 2016 was approximately 150 days.

Best practice suggests that in the immediate aftermath of an attack, firms must:

  • Understand how the breach occurred in order to immediately remediate any deficiency that was exploited

  • Quickly identify where the firm’s exposure is, especially which clients are impacted

  • Recover lost data

  • Notify the relevant regulators and the ICO of the breach. This will become even more pertinent with the new EU General Data Protection Regulation that comes into force next year

A document management system (DMS) plays a key role here – after all, this is where the firm’s ‘crown jewels’ reside and so protecting the sensitive information within it is essential. Today, technology has advanced and security is no longer simply about complex passwords. Adoption of analytics and machine learning is essential to proactive data security and loss mitigation.

Firms that deploy iManage Work, a next generation DMS, are now able to automatically profile the typical behaviour of users and use this information to identify potential social engineering attacks and take pre-emptive action. This is provided by the recently announced Threat Manager module, which uses machine learning to automatically generate a digital ‘finger print’ of all employees based on their individual history of interaction with data in the DMS. Any deviation from this baseline is a good indication of anomalous activity and warrants investigation. The behavioural analytics used by Threat Manager is significantly more advanced than simple threshold reporting; and by understanding the matters, clients and practice areas relevant to the firm’s personnel, it can identify a hacker using stolen credentials to view even a small number of files as being an anomaly, and flag this for immediate action. In fact, an interesting finding during the testing phase of Threat Manager was that it is also able to predict with high accuracy, personnel who are likely to be leaving the firm. This is often a concern for firms due to potential loss of organisational and client IP.

In addition to threat detection, Threat Manager also provides capabilities to assist with investigating the timeline of the attack, leveraging the comprehensive audit trails within the DMS to identify precisely which matters and documents have been accessed. The tool is simple to use and means this crucial capability can be removed from the IT department and given to a more appropriate function such as the Risk and Compliance team. Those responsible regularly receive reports on alerts detected with corresponding scores based on the threat level, enabling them to take immediate action to prevent any further wrongdoing.

Given the nature of the threats faced, adoption of behavioural analytics and machine learning is going to be a key tool for firms wanting to take a proactive stance on data security and providing the assurance their clients are increasingly looking for.

Note: As an iManage partner, the new Threat Manager module for iManage Work is available from Ascertus. Please get in touch via for more information on the module and/or a demonstration of the product.

About Frank White

Frank has worked in legal IT for over 25 years, having occupied a number of technical roles through to IT Director of Ince & Co, responsible for delivering a professional yet personalised IT service and cost-appropriate IT solutions to a diverse global practice. Frank joined iManage in February 2015 as a Subject Matter Expert to help iManage customers realise the best return on their investment in their iManage products, and to enable the company to deliver better products and create more value.

Thursday, March 16, 2017

The Difference Between Metrics and KPIs

By Dr. Michael Tal, Managing Director, BusyLamp

The biggest challenges faced by legal departments according to corporate counsel are improving operational efficiency, ensuring regulatory compliance, alleviating budget pressure, delivering despite a staff shortage and making better use of technology. So, in a nutshell, doing ‘more with less’ is an overriding objective. Often, corporate counsel often has a gut feeling that everyone in the department is “doing more with less”, but struggle to quantify the “more” and the “less”.

All the above-mentioned challenges could to a large extent be reduced by using legal analytics. Existing data can help to measure these items and drive quantifiable improvements. In addition, it allows the legal department to speak the language of company executives and senior management – i.e. the language of metrics and Key Performance Indicators (KPIs).

To obtain high-quality, trustworthy analytics, one must feed it with good data. There are several factors that determine quality data. With regard to legal spend management, these are for example: categorised matters, Uniform Task Based Management Systems (UTBMS) task-coded line items, robust line item descriptions and line items that are not block-billed.

Law departments have an abundance of available data, like historic bills from their outside counsel, which contain a wealth of valuable information, if they could mine it. This requires a structure and using legal analytics platforms is a good idea. These platforms can structure, aggregate, organise and compare data in meaningful ways to make it truly actionable in the form of metrics and KPIs.

To speak bluntly, metrics in organisations tend to be what can be easily measured. So often decision makers discount the metrics that don’t appear to achieve the outcome they desire at the least possible cost. Nevertheless, in order to convey the resounding impact of process innovation and improvement through the use of applied metrics on the legal department, it’s important to define success factors, limitations and choose KPIs that align with the established strategy. This makes the relationship between a KPI and a metric tangible – a KPI is simply a metric that is tied to a target.

A metric is generally understood to be a standard of measurement and that can be used as a way of quantitatively assessing the efficiency of performance with respect to a particular process. In order for a metric to be useful, however, there must be full agreement in advance as to how the relevant data will be collected, organised and displayed. Metrics cannot be used in isolation and their results must be evaluated against pre-agreed standards. This exercise is referred to as “benchmarking”, post which a large number of “performance indicators” can be identified. The focus must be on those that are truly “key” – meaning that KPIs should have a sensible and sustainable architecture, which aligns with the company’s overall strategy and are easily understood by stakeholders while being limited to those areas that can create the most value for the organisation.

There is no “right answer” for what metrics a legal department should use and the KPI programme must be implemented, but there are some questions that can help drive the decision-making process:

  1. Why are the KPIs being developed? For example, is it to demonstrate the value of the legal department; or just track trends?
  2. What are the department’s goals? What are the company’s overarching goals? Is there a strategy you could review?
  3. What aspects of performance are of value to leadership (company, law department, business units)?
  4. Does anyone have extensive experience of developing and implementing KPIs within the organisation?
  5. Are metrics being tracked in other areas? What do those look like and what is the process for evaluating and measuring them? What were the lessons learned?
  6. What processes and tools are in place already that could help with measurement? What processes and tools would need to be implemented and what does that effort look like?
  7. Where does the KPI programme fit with the priorities of the department?

When determined following a diligent process, metrics can help incentivise the legal department and they can be embedded within the workflow. KPIs, on the other hand, should be used to establish goals and objectives for further improvement. Note however, that the goals should not simply be to achieve a “better score”, but rather be supported by specific “next steps” that can be understood, implemented and measured. This approach will enable the department to truly strive for and achieve definite improvements.