SE Labs

Posts tagged 'email'

Email security: Is it any good against hackers?

World’s first in-depth, public test of security services vs. targeted attacks.

This email security test report is the product of two years of advanced threat research. We have worked with the security companies themselves and with their customers.  We have monitored what the bad guys have been doing and identified and replicated real-world email threats that affect everyone generally, and also specific types of businesses.

There is no report like this anywhere in the public domain. We are extremely proud to present the results here.

Read more >

SE Labs Annual Report 2019

SE Labs has been working at the core of the cyber security industry since its launch in 2016. We work with all of the major developers of IT security products as well as their main customers and even investors looking to increase their chances when betting on emerging technologies.

Read more >

How well do email security gateways protect against targeted attacks?

Email security test explores how and when services detect and stop threats.

Latest report now online.

This new email protection test shows a wide variation in the abilities of the services that we have assessed.

You might see the figures as being disappointing. Surely Microsoft Office 365 can’t be that bad? An eight per cent accuracy rating seems incredible.

Literally not credible. If it misses most threats then organisations relying on it for email security would be hacked to death (not literally).

But our results are subtler than just reflecting detection rates and it’s worth understanding exactly what we’re testing here to get the most value from the data. We’re not testing these services with live streams of real emails, in which massive percentages of messages are legitimate or basic spam. Depending on who you talk to, around 50 per cent of all email is spam. We don’t test anti-spam at all, in fact, but just the small percentage of email that comprises targeted attacks.

In other words, these results show what can happen when attackers apply themselves to specific targets. They do not reflect a “day in the life” of an average user’s email inbox.

We have also included some ‘commodity’ email threats, though – the kind of generic phishing and social engineering attacks that affect everyone. All services ought to stop every one of these. Similarly, we included some clean emails to ensure that the services were not too aggressively configured. All services ought to allow all these through to the inbox.

So when you see results that appear to be surprising, remember that we’re testing some very specific types of attacks that happen in real life, but not in vast numbers comparable to spam or more general threats.

The way that services handle threats are varied and effective to greater or lesser degrees. To best reflect how useful their responses are, we have a rating system that accounts for their different approaches. Essentially, services that keep threats as far as possible from users will win more points than those who let the message appear in or near the inbox. Conversely, those that allow the most legitimate messages through to the inbox rate higher than those which block them without the possibility of recovery from a junk folder or quarantine.

If you spot a detail in this report that you don’t understand, or would like to discuss, please contact us via our Twitter or Facebook accounts.

SE Labs uses current threat intelligence to make our tests as realistic as possible. To learn more about how we test, how we define ‘threat intelligence’ and how we use it to improve our tests please visit our website and follow us on Twitter.
Our latest reports, for enterprisesmall business and home users are now available for free from our website. Please download them and follow us on Twitter and/or Facebook to receive updates and future reports.

Tough test for email security services

essp-4202383

Our latest email cloud security test really challenged the services under evaluation.

Latest report now online.

Last summer we launched our first email cloud security test and, while it was very well received by our readers and the security industry as a whole, we felt that there was still work to do on the methodology.

This report shows the results of six months of further development, and a much clearer variation in the capabilities of the services under test.

The most significant change to the way we conducted this test lies in the selection of threats we used to challenge the security services: we increased the number and broadened the sophistication.

Whereas we might have used one fake FBI blackmail email previously, in this test we sent 10, each created using a different level of sophistication. Maybe a service will detect the easier versions but allow more convincing examples through to the inbox?

We wanted to test the breaking point.

We also used a much larger number of targeted attacks. There was one group of public ‘commodity’ attacks, such as anyone on the internet might receive at random, but also three categories of crafted, targeted attacks including phishing, social engineering (e.g. fraud) and targeted malware (e.g. malicious PDFs).

Each individual attack was recreated 10 times in subtly different but important ways.

Attackers have a range of capabilities, from poor to extremely advanced. We used our “zero to Neo” approach to include basic, medium, advanced and very advanced threats to see what would be detected, stopped or allowed through.

The result was an incredibly tough test.

We believe that a security product that misses a threat should face significant penalties, while blocking legitimate activity is even more serious.

If you’re paying for protection threats should be stopped and your computing experience shouldn’t be hindered. As such, services that allowed threats through, and blocked legitimate messages, faced severe reductions to their accuracy ratings and, subsequently, their chances of winning an award.

Intelligence-Led Testing

We pay close attention to how criminals attempt to attack victims over email. The video below shows a typically convincing attack that starts with a text message and ends stealing enough information to clean out a bank account.
SE Labs uses current threat intelligence to make our tests as realistic as possible. To learn more about how we test, how we define ‘threat intelligence’ and how we use it to improve our tests please visit our website and follow us on Twitter.

Anatomy of a Phishing Attack

phishing_magnifying_glass_fi-3673555Who attacked a couple of Internet pressure groups earlier this year? Jon Thompson examines the evidence.

For those on those of us engaged in constructing carefully-crafted tests against client email filtering services, the public details of an unusually high-quality spear-phishing attack against a low value target make for interesting reading.

In this case, there were two targets: Free Press, and Fight for the Future. The attack, dubbed “Phish for the Future” in a brief analysis by the Electronic Frontier Foundation, is curious for several reasons.

Free Press is a pressure group campaigning for an open internet, fighting media consolidation by large corporations, and defending press freedom. Fight for the Future works to protect people’s basic online freedoms. Objectively, they’re working for a better online future, which makes the whole affair stand out like a pork buffet at a bar mitzvah.

The first thing that struck me was that the emails were apparently all sent during office hours. The time zones place the senders anywhere between Finland and India, but apparently resolve to office hours when normalised to a single zone.

Another interesting aspect is that even though the emails were sent on 23 active days, the attackers didn’t work weekends. This immediately marks them out as unusual. Anyone who’s run an email honeypot knows that commodity spam flows 24 hours a day.

The attackers first tried generic phishing expeditions, but quickly cranked up their targeting and psychological manipulation. This begs an interesting question: If you’re an experienced, professional, disciplined crew, why jeopardise the operation by beginning with less convincing samples that may alert the target to be on the lookout? Why didn’t they simply start with the good stuff, get the job done, and move on?

One possible explanation is that the attackers were trainees on a course, authorised to undertake a carefully controlled “live fire” exercise. Psychologically manipulative techniques such as pretending to be a target’s husband sending family photos, or a fan checking a URL to someone’s music, imply a level of confident duplicity normally associated with spying scandals.

The level of sophistication and persistence on display forms a shibboleth. It looks and smells somehow “wrong”. The published report reveals an attention to detail and target reconnaissance usually reserved for high value commercial targets. Either the attackers learn at a tremendous rate
through sheer interest alone, or they’re methodically being taught increasingly sophisticated techniques to a timetable. If it was part of a course, then maybe the times the emails were sent show a break for morning coffee, lunch and afternoon tea, or fall into patterns of tuition followed by practical exercises.

phishing2b-6448783The timing of the complete attack also stands out. It began on 7th July, ended on 8th August, and straddled the Net Neutrality Day of Action (12th July). With a lot happening at both targets during that time, and one assumes a lot of email flying about, perhaps the attackers believed they stood a better chance when the staff were busiest.

So, to recap, it looks like highly motivated yet disciplined attackers were operating with uncommonly sophisticated confidence against two small online freedom groups. Neither target has the business acumen of a large corporation, which rules out criminal gain, and yet an awful lot of effort was ranged against them.

The product of phishing is access, either to abuse directly or to be sold to others. Who would want secret access to organisations campaigning for online freedom? Both targets exist to change minds and therefore policy, which makes them political. They’re interesting not only to governments, but also to media companies seeking to control the internet.

I’m speculating wildly, of course. The whole thing could very easily have been perpetrated by an under-worked individual at a large company, using their office computer and keeping regular hours to avoid suspicion. The rest is down to ingenuity and personal motivation.

We’ll never know the truth, but the supporting infrastructure detailed in the EFF report certainly points to some considerable effort over a long period of time. If it was an individual, he’s out there, he’ll strike again, and he learns fast. In many ways, I’d prefer it to have been a security service training new recruits.

Email hosted protection tested

emailhostedprotection-5850764

Our first cloud-based email protection report is now available.

Email provides a route right into the heart of our computers, phones and other devices. As such, it is frequently abused to perform a variety of attacks against potential victims of cybercrime.

Latest report now online.

The sophistication of attacks vary but many rely on our almost unbreakable instinct to open, read and interact with messages sent to work and personal email accounts. Businesses rely on email security services to filter out large numbers of such attacks.

The range of attack types in the real world is wide, but in general we consider there to be two main categories: targeted attacks, in which the attacker attempts to target a specific individual; and public attacks, which spread wide and far in an attempt to compromise as many people as possible.
Many of the same techniques are used in public and targeted attacks. The least technically sophisticated include requests for a money transfer or banking login credentials. More credible attempts include professionally-formatted emails and links to fake websites designed to trick users into entering their valuable details.

Attackers with more resources may use malware to achieve their goals, either in the form of attached files or by linking to websites that exploit visiting computers.
SE Labs monitors email threats in real-time, analysing large  numbers of messages and extracting samples that represent  large groups of those threats. Human testers then manually verify that any malware included works properly before re-sending these threats to our own accounts through the tested services.

We also generate targeted attacks using the same tools and techniques used by advanced attackers. In gathering threats this way we achieve a realistic and relevant coverage of existing threats in a small set of test samples.

Our latest reports, for enterprise, small business and home users are now available for free from our website. Please download them and follow us on Twitter and/or Facebook to receive updates and future reports.

Brexit and Cybersecurity

Is the UK headed for a cybersecurity disaster?

istock-big-ben-parliament-standard-5154835

With Brexit looming and cybercrime booming, the UK can’t afford major IT disasters, but history says they’re inevitable.

The recent WannaCry ransomware tsunami was big news in the UK. However, it was incorrectly reported that the government had scrapped a deal with Microsoft to provide extended support for Windows XP that would have protected ageing NHS computers. The truth is far more mundane.

In 2014, the government signed a one-year deal with Microsoft to provide security updates to NHS Windows XP machines. This was supposed to force users to move to the latest version of Windows within 12 months, but with a “complete aversion to central command and control” within the NHS, and no spare cash for such an upgrade, the move was never completed.

This isn’t the first IT Whitehall IT disaster by a very long way.

During the 1990s, for example, it was realised that the IT systems underpinning the UK’s Magistrates’ Courts were inadequate. It was proposed that a new, unified system should replace them. In 1998, the Labour government signed a deal with ICL to develop Project Libra. Costing £146m, this would manage the courts and link to other official systems, such as the DVLA and prisons systems.

Described in 2003 as “One of the worst IT projects ever seen“, Project Libra’s costs nearly tripled to £390m, with ICL’s parent company, Fujitsu, twice threatening to pull out of the project.

This wasn’t Labour’s only IT project failure. In total, it’s reckoned that by the time the government fell in 2010, it had consumed around £26b of taxpayer’s money on failed, late and cancelled IT projects.

The coalition government that followed fared no better. £150m paid to Raytheon in compensation for cancelling the e-Borders project, £100m spent on a failed archiving system at the BBC, £56m spent on a Ministry of Justice system that was cancelled after someone realised there was already a system doing the same thing: these are just a few of the failed IT projects since Labour left office seven years ago.

The Gartner group has analysed why government IT projects fail, and discovered several main factors. Prominent amongst these is that politicians like to stamp their authority on the nation with grandiose schemes. Gartner says such large projects fail because of their scope. It also says failure lies in trying to re-implement complex, existing processes rather than seeking to simplify and improve on them by design. The problem is, with Brexit looming, large, complex systems designed to quickly replace existing systems are exactly what’s required.

ukba_and_police-7387838

A good example is the ageing HM Customs & Excise CHIEF system. Because goods currently enjoy freedom of movement within the EU, there are only around 60 million packages that need checking in through CHIEF each year. The current system is about 25 years old and just about copes. Leaving the EU will mean processing an estimated 390 million packages per year. However, the replacement system is already rated as “Amber/Red” by the government’s own Infrastructure and Projects Authority, meaning it is already at risk of failure before it’s even delivered.

Another key system for the UK is the EU’s Schengen Information System (SIS-II). This provides real time information about individuals of interest, such as those with European Arrest Warrants against them, terrorist suspects, returning foreign fighters, missing persons, drug traffickers, etc.

Access to SIS-II is limited to countries that abide by EU European Court of Justice rulings. Described by ex-Liberal Democrat leader Nick Clegg as a “fantastically useful weapon” against terrorism, after Brexit, access to SIS-II may be withdrawn.

Late last year, a Commons Select Committee published a report identifying the risks to policing if the UK loses access to SIS-II and related EU systems. The report claimed that then-Home Secretary Theresa May had said that such systems were vital to, “stop foreign criminals from coming to Britain, deal with European fighters coming back from Syria, stop British criminals evading justice abroad, prevent foreign criminals evading justice by hiding here, and get foreign criminals out of our prisons.

The UK will either somehow have to re-negotiate access to these systems, or somehow quickly and securely duplicate them and their content on UK soil. To do so, we will have to navigate the EU’s labyrinthine data protection laws and sharing agreements to access relevant data.

If the UK government can find a way to prevent these and other IT projects running into problems during development, there’s still the problem of cybercrime and cyberwarfare. Luckily, there’s a strategy covering this.

In November 2016, the government launched its National Cyber Security Strategy. Tucked in amongst areas covering online business and national defence, section 5.3 covers protecting government systems. This acknowledges that government networks are complex, and contain systems that are badly in need of modernisation. It asserts that in future there will be, “no unmanaged risks from legacy systems and unsupported software”.

The recent NHS WannaCry crisis was probably caused by someone unknowingly detonating an infected email attachment. The Strategy recognises that most attacks have a human element. It says the government will “ensure that everyone who works in government has a sound awareness of cyber risk”. Specifically, the Strategy says that health and care systems pose unique threats to national security due to the sector employing 1.6 million people in 40,000 organisations.

The problem is, the current Prime Minister called a snap General Election in May, potentially throwing the future of the Strategy into doubt. If the Conservatives maintain power, there’s likely to be a cabinet reshuffle, with an attendant shift in priorities and funding.

european-union-flag-std_1-9767927

If Labour gains power, things are even less clear. Its manifesto makes little mention of cyber security, but says it will order a complete strategic defence and security review “including cyber warfare”, which will take time to formulate and agree with stakeholders. It also says Labour will introduce a cyber charter for companies working with the Ministry of Defence.

Regardless of who takes power in the UK this month, time is running out. The pressure to deliver large and complex systems to cover the shortfall left by Brexit will be immense. Such systems need to be delivered on time, within budget and above all they must be secure – both from internal and external threats.

Back from the Dead

email-1932571Forgotten web sites can haunt users with malware.

Last night, I received a malicious email. The problem is, it was sent to an account I use to register for web sites and nothing else.

Over the years, I’ve signed up for hundreds of sites using this account, from news to garden centres. One of them has been compromised. The mere act of receiving the email immediately marked it out as dodgy.

The friendly, well written message was a refreshing change from the usual approach, which most often demands immediate, unthinking action. The sender, however, could only call me “J” as he didn’t have my forename. There was a protected file attached, but the sender had supplied the password. It was a contract, he said, and he looked forward to hearing back from me.

The headers said the email came from a French telecoms company. Was someone on a spending spree with my money? My PayPal and bank accounts showed no withdrawals.

Curious about the payload, I spun up a suitably isolated Windows 10 victim system, and detonated the attachment. It had the cheek to complain about having no route to the outside world. I tried again, this time with an open internet connection. A randomly-named process quickly opened and closed, while the file reported a corruption. Maybe the victim system had the wrong version of Windows installed, or the wrong vulnerabilities exposed. Maybe my IP address was in the wrong territory. Maybe (and this is more likely) the file spotted the monitoring software watching its every move, and aborted its run with a suitably misleading message.

Disappointed, after deleting the victim system I wondered which site out of hundreds could have been compromised. I’ll probably never know, but it does reveal a deeper worry about life online.

Over the years, we all sign up for plenty of sites about which we subsequently forget, and usually with whichever email address is most convenient. It’s surely only a matter of time before old, forgotten sites get hacked and return to haunt us with something more focused than malicious commodity spam – especially if we’ve been silly enough to provide a full or real name and address. Because of this, it pays to set up dedicated accounts for registrations, or use temporary addresses from places such as Guerrilla Mail.

17 Things Spammers Get Wrong


No one publishes successful phishing and ransomware emails. Jon Thompson thinks he knows why.

ransomware-8145580The headlines say phishing scams are at an all-time high, and ransomware is growing exponentially, but conspicuous by their absence are examples of the emails behind successful attacks. It’s becoming the cliché in the room, but there may be a reason: embarrassment.

Running an email honeypot network, you receive a flood of malicious email every day. Most is littered with glaring errors that point to lazy, inarticulate crooks trying to make the quickest buck from the least effort. When you do come across a rare, well though-out campaign, it shines like a jewel in a sea of criminal mediocrity.

To the average spammer, however, it’s all just a numbers game. He cranks the handle on the botnet, so to speak, and money comes out.

This poses an important question: why, given the quality of most malicious spam, are new ransomware infections and high profile phishing attacks still making headlines almost every single day? Clearly, we’re massively overestimating the amount of effort and intelligence invested by spammers.

With that in mind, what follows is a short list of 17 mistakes I routinely see, all of which immediately guarantee that an email is malicious. There are others, but these are the main ones. If this list reflects the mistakes found in the spam behind the headlines, then the size yet lack of sophistication of the problem should become apparent.

1.    No Subject Header

This error is particularly prevalent in ransomware campaigns. Messages whose payloads have very low VirusTotal scores are being sent with no subject header. Maybe the sender thinks it’ll pique the curiosity of the recipient, but it should also alert spam filters even before they examine the attachment.

2.    No Set Dressing

tesco-6478043

Look at any real communication from a bank, PayPal, a store, etc. It is well formatted, the HTML is clean, the language is clear, and the branding is obvious. Legitimate companies and banks don’t tend to send important messages in plain text.

3.    Generic Companies

generic-1081819

Generic companies are rare but I do occasionally see them. Who is “the other financial institution” and why has it refused my transaction? Vague, instantiated company names like this, with an accompanying attachment, are clear indicators of spam.

4.    Multiple Recipients

This is another example of laziness on the part of spammers. OK, they may have found an open relay to willingly spread messages rather than buy extra time on a botnet, but anything other than a one-to-one sender to recipient ratio should be an instant red flag.

5.    Poor Salutation

Much apparently personalised spam doesn’t use a competent salutation, or uses a salutation that is simply the user name part of the email address (i.e.: “Dear fred.smith”). It would take effort to code a script that personalises the messages by stripping off the first name and capitalising the initial. Effort is the enemy of the fast buck.

6.    No Body Text

Sending an email with a tantalizing subject header such as “Overdue – Please Respond!” but no body text explaining what or why it’s overdue is as common in commodity ransomware as having no subject header. The attack again relies entirely on the natural curiosity of the recipient, who can and should simply ignore it. Spam filters should also take a keen interest.

7.    Auto-translated Body Text

paypal2-9354648

Machine translation has the amusing habit of mapping the grammar of one language onto another, resulting in errors that no native speaker would ever make. Manual translation by a highly fluent speaker is far superior to machine translation, but the translator must also have knowledge of the subject matter for his text to appear convincing. Again, this is effort.

8.    The Third Person

This is a great example of a spam writer trying to distance himself from his crime. “PayPal has detected an anomaly in your account” and “they require you to log in to verify your account” just look weird in the context of a security challenge. This is supposed to be from PayPal, isn’t it?

9.    Finger Trouble

apple2bicloud2b2-6952704

I’m fast concluding that some cybercriminals really do wear thick leather gloves while typing, just like in the pictures. Either that or they’re blind drunk. Random punctuation marks and extra characters that look like they’ve been hit at the same time as the correct ones don’t make a good impression. Simply rejecting emails that have more than a certain percentage of spelling mistakes might prevent many of these messages from getting through.

10.    Unexpected Plurals and Tenses

Using “informations” instead of “information” is a dead giveaway for spam and should be blocked when in combination with other indicators. Phrases such as “we detect a problem” instead of “we detected a problem” also stick out a mile.

11.    Missing Definite Article

Many spam emails stand out as somehow “wrong” because they miss out the definite article. One recent example I saw read: “Access is blocked because we detect credit card linked to your PayPal account has expired.” An associated Yandex.ru return address gave the whole thing a distinct whiff of vodka.

12.    The Wrong Word

“Please review the document and revert back to us immediately”. Revert? Really? Surely, you mean “get back”, not “revert back”. It may be difficult for spam filters to weed out this kind of error, but humans should spot it without difficulty.

13.    Misplaced Emphasis

paypal-1531640

Unusually capitalised phrases such as “You must update Your details to prevent Your Account from being Suspended” look weird. Initial capitalisation isn’t used for emphasis in English sentences, and hints at someone trying to make the message sound more official and urgent than it is.

14.    Tautological Terrors

cps-7972121

“It is extremely mandatory that you respond immediately”. Not just mandatory but extremely mandatory? Wow, I’d better click that link right away! Urgent calls to action like this overplay the importance of the message in ways that mark them out as fake.

15.    Grandiosity

splayer-6973270

Using grand words where normal ones should appear to make a message sound more authoritative are a dead giveaway.  Here’s an example from last September when a gang famously tried to distribute malware on the back of a new media player release: “To solemnise the release of our new software”. Solemnise means to mark with a formal ceremony.

What they really meant was: “To mark the release of our new software”.  The whole message was also riddled with the most outrageous auto-translate errors that it made difficult reading.

16.    Overly-grand Titles

Why would the Microsoft Chief Support Manager be contacting me personally all the way from the US to give me a refund? Wouldn’t he delegate this important work to a local minion? Similarly, the head of the IMF doesn’t usually spend their days emailing strangers about ATM cards stacked high with cash.

17.    Obfuscated URLs

If the collar doesn’t match the cuffs, it’s a lie. In other words, if the message contains the name of a high-street bank (for example) and a URL from a shortening service such as bit.ly, spam filters should be blocking the message without question, regardless of the rest of the content.

About

SE Labs Ltd is a private, independently-owned and run testing company that assesses security products and services. The main laboratory is located in Wimbledon, South London. It has excellent local and international travel connections. The lab is open for prearranged client visits.

Contact

SE Labs Ltd
Hill Place House
55A High Street
Wimbledon
SW19 5BA

020 3875 5000

info@selabs.uk

Press