The security risks of modern Voice over Internet Protocol
Cyber crime is becoming an increasingly significant and growing global problem that affects all sectors with an on-line platform or service. The UK Cyber Security… Read more
Cyber crime is becoming an increasingly significant and growing global problem that affects all sectors with an on-line platform or service. The UK Cyber Security Breaches Survey 2016 commissioned by the Department for Culture, Media and Sport (DCMS) found that 25% of companies experience a cyber-breach at least once a month. At the annual Black Hat USA cyber security conference, warnings were delivered about the security inadequacies of modern VOIP (Voice over Internet Protocol) or unified communications systems which are sometimes overlooked when corporates or individuals assess their cyber-attack vulnerabilities. The incorporation of VOIP into a corporate network means that VOIP tends to be another service running over the IP network and is therefore another door through which hackers can gain access to a wider system or underlying infrastructure.
What is VOIP?
VOIP or Voice over Internet Protocol is a service that is used to send voice transmission over a broadband internet connection. It transmits analogue voice signals as digital packets over the internet instead of using the traditional public switched telephone network. This convergence of data and voice means that VOIP is another data service over the IP network and there is no physical separation of the networks. For VOIP providers it, it is an electronic communications service and is regulated in the UK by Ofcom. Since it was first launched, gateways have been subsequently created as an add-on to VOIP to allow computer to telephone calls and vice versa. Corporate VOIP services can now also be integrated with other corporate enterprise systems to search for users in internal corporate directories and deliver voicemail to a desktop.
The advantage of VOIP is that standard call charges are not incurred. However, the disadvantage is that as VOIP is delivered using a third party application, the same requirements to patch and update the software apply. An additional issue is that the standard firewalls cannot always analyse the data carried by the SIP Protocol (the protocol commonly used for controlling VOIP calls) and consequently cannot determine whether a call is legitimate or fraudulent.
There has been a reported increase in cyber attacks on VOIP infrastructure and particularly on UK servers with most attacks taking place outside of regular working hours.
The VOIP Security Alliance has identified the following five types of threats to a VOIP service:
- Social threats: misrepresentation of identity, authority, rights or content – it is possible to doctor voices or change incoming phone numbers.
- Interception: eavesdropping on calls (which without lawful authority is illegal in the UK)
- Service abuse: hacking into a system to make calls to premium rate numbers or international numbers (one of the most common attacks).
- Intentional interruption of service: a denial of service attack particularly targeted to the VOIP system.
- Other physical interruptions: such as loss of power to the servers hence bringing the service down.
From a practical perspective, companies often assume that the VOIP service should be handled by a telephony specialist when in fact, the security of the system is a networking matter. Obviously, securing a PBX (Private Branch Exchange) completely from the outside world would render the service totally redundant but it is a case of focusing on minimising vulnerabilities. This can be from a practical perspective through mailbox passwords, implementing security features provided with the service such as barring access to premium rate numbers and using a VPN to carry the traffic between a remote endpoint and the PBX.
Information and Cyber Security Legal Position
The current general legal position in the UK requires those organisations that control personal data to take appropriate technical and organisational measures against unauthorised or unlawful processing of personal data and against accidental loss or destruction of or damage to personal data (Data Protection Principle 7, Data Protection Act 1998).
In practice this requires corporates to:
- Adopt security measures (both physical and technological) commensurate with the type of data being held and the harm that may result from a security breach;
- Be ready to respond to any breach of security swiftly and effectively; and
- Be clear as to who is responsible for information security within the organisation and back this up with policies, procedures and well trained staff in relation to the security measures, reporting and responsibility.
In essence, the VOIP service and infrastructure should form part of the IT and Communications estate that has equally as stringent security measures applied to it (yet adapted to address the protocols used to provide the service).
The UK Privacy and Electronic Communication Regulations also need to be complied with for VOIP service providers. Essentially the technical and organisational measures to be adopted mirror those in the Data Protection Act with the following additional requirements that they must:
- Ensure that personal data can be accessed only by authorised personnel for legally authorised purposes;
- Ensure the implementation of a security policy with respect to the processing of personal data; and
- Notify the Information Commissioner (ICO) in the event of a personal data breach within 24 hours of becoming aware of the basic facts, with full details as soon as possible (with a more detailed follow up notification 72 hours later if applicable) and notify the subscriber without undue delay if the personal data breach is likely to adversely affect the personal data or privacy of a subscriber or user;
- A log of any breaches must be kept and should be submitted to the ICO on a monthly basis.
Any notification to a user of the system is not required if the service does not have a direct relationship with the end user (but it must notify the organisation that does). Alternatively, if the service provider has demonstrated to the ICO’s satisfaction that it has implemented measures that render the data unintelligible to any person not authorised to access it and that those measures were applied to the data concerned in that breach. The ICO also has an audit right to assess compliance with the notification requirements and a failure by a service provider to comply with the notification requirements may attract a fine of £1000. Where there remains significant risk to the security of the service, the service provider shall inform the subscribers concerned of the nature of the risk, any appropriate measures the subscriber may take to safeguard against that risk and the likely cost of such measures.
Regardless of Brexit, the general consensus seems to be that the EU General Data Protection Regulation (GDPR) will come into force in May 2018 and that the UK will adopt legislation that closely mirrors both that and the recently adopted Network and Information Security Directive once the UK leaves the European Union. Obviously, any EU laws already in effect in the UK will need to be revoked or repealed before they cease to be law. With this in mind, the data breach notifications that currently apply to public electronic communications service or network providers i.e. the VOIP service providers will now also fall on those using the services in the event of a data breach, with the potential for significant fines if this timescale is not met (assuming fines are at a similar level to those set in the GDPR). The NIS Directive’s application is broader than incidents affecting personal data – although it will only apply to ‘essential service providers’ determined as such by the Government and digital service providers such as search engines, cloud computing providers and online market places. There are notification requirements where there is an incident to the network and information systems of that service would have significant disruptive effect on the provision. This could include the failure of a VOIP or unified communications service whether or not there has been a breach of personal data.
For those faced with procuring a VOIP service, it appears to be an increasingly common position from VOIP providers that they are a more conduit for the data and do not store personal data. Once the GDPR applies, the distinction between a processor and controller will be less stark as processors will also incur liability. In addition, companies should ideally place and enforce their own information security requirements on the service provider and if this is not possible, ensure that the service providers own information security requirements are sufficient. It is always worth carrying out due diligence and asking if there have been data breaches or security issues arising from the service. The question of who incurs the cost in the event there has been fraudulent use of the system is always relevant – particularly when it becomes an issue of whether the fraud has occurred due to the security vulnerabilities of a particular VOIP service. Such an issue is generally always contentious and where it Is a risk that the VOIP service provider is not willing to assume this should also be built into any business case of assessing the cost savings of moving to the service. In the event that a main switchboard and/or phone lines (internal and external) may cause significant disruption, the limits of liability should be negotiated with the size and nature of the business and service credits in mind.
The costs savings brought about by migrating to VOIP systems are clear but the convergence of the networks and the increased cyber vulnerabilities this brings about as companies add a service which by its very nature must be connected to the outside world should not be overlooked as cybersecurity increasingly becomes a Board level issue where reporting on security breaches will in May 2018 be a matter for all companies not just the telcos and with potentially significant fines attached.
This article was previously published in the September 2016 issue of Cyber Security Practitioner.