The UK National Cyber Security Centre has issued a stark warning of the dangers posed by an ongoing Russian campaign of cyber espionage and attacks against enterprise systems and cloud environments in Europe and around the world.

The Cyber Security Advisory is a joint creation of the NCSC working with the NSA, CISA and FBI in the United States and it details the malicious cyber activity of Russia’s military, the Russian General Staff Main Intelligence Directorate (GRU) and the 85^th Main Special Service Center (GTsSS), against organisations around the world.  The campaign has been underway since mid 2019 and has targeted: government and military organisations, defence contractors, energy companies, higher education, logistics, law firms, media, political consultants or political parties and think tanks.

The report titled: “Russian GRU Conducting Global Brute Force Campaign to Compromise Enterprise and Cloud Environments” explains the tactics, techniques and procedures (TTPs) used by the threat actors to gain access to targeted networks, compromise credentials and move laterally within the network and gather and steal data.  The activities of these Russian military groups are also referred to as: Fancy Bear, APT28 and Strontium in the tech press.

This attack is mainly a brute force attack which uses a Kubernetes cluster to power a widespread brute force attack against Microsoft Office 365 cloud environments, and other hosted and on premises systems.  Once access has been secured, the attackers are either taking advantage of the access themselves or passing the credentials to another team to facilitate their own objectives.

What is a brute force attack?

A brute force attack is an attempt to gain access to a system by repeatedly trying different password and or username combinations until a valid one is found.

There are several variations used by attackers, the main ones being credential stuffing and password spraying.

Credential Stuffing uses automated tools to test a large volume of stolen usernames and passwords across multiple sites until one works.  It is helped when people re-use passwords on multiple accounts, so a password revealed in a breach on site A, turns out to be a valid login on site B for example.

Password spraying on the other hand attempts to access a large number of sites using either known or easily guessed (think ‘admin’) usernames with a smaller list of commonly used passwords.

How to defend against brute force attacks

There are several steps security managers can take to protect their networks against credential theft attacks in general, as well as this Russian campaign:

Multi-factor Authentication

The best and first step to defend against credential theft or compromise is to use multi-factor authentication (MFA) – which is widely available on most cloud services and can be readily added to Active Directory for on premise systems.  Strong authentication factors (such as One Time Passwords, smart cards or tokens) require something more than just the password to access an account and so protect the network against compromise of the password.

Time based defence

Brute force attacks can be effective because they attempt many combinations in a relatively short period of time.  By increasing the time taken to perform an attack, the effectiveness of brute force can be reduced.  Consider implementing some of the following time based defences:

Time-out protection increases the time between each login attempt by slowing the user interface.  This could be implemented within the login page itself – and so protecting the authentication service from having to deal with the attempts, or the authentication service itself could manage the throttling of the response speed.  As soon as the time taken between failed login attempts increases beyond a few seconds most brute force attacks become infeasible.

Lock-out protection imposes a hard limit on the number of failed login attempts permitted within a defined window. For example, if more than 10 login attempts fail within a 24 hour period, the account in question becomes locked and human intervention is required to unlock the account (and investigate why the lockout happened).  Care should be taken with this approach to ensure a malicious user could not deliberately cause an account to become locked that is required for normal system operations – such as a service owner account – and so cause a component to fail resulting in a kind of denial-of-service event.

Dictionary and complexity defence

When a password is first set or changed, it can be checked against a dictionary of disallowed words (common password choices) and its complexity evaluated.  This prevents passwords from being chosen that will be vulnerable to brute force attack – either because they are common and easily guessed or too short.

Use captchas for people

For human interfaces, use a captcha system to hinder scripted attacks.

Check access logs

Use automated tools, such as SIEM, to analyse audit access logs and detect anomalous access requests and alert security staff to investigate.

Adopt Zero Trust Security

Move towards a Zero Trust Security model that considers additional attributes when making authorisation and authentication decisions (such as device information, access path and environmental factors) allowing requests using valid passwords to be denied.  One first step in this direction recommended in the NCSC report is to automatically deny connections attempts made through anonymisation services such as the TOR network or VPN services not normally used by the business.