STE WILLIAMS

Anomaly in pen-test tool made malware servers visible

For four years, a security company was able to track command and control (CC) traffic generated by several well-known hacking groups thanks to a tiny anomaly in a penetration-testing tool.

This news emerged in a write-up by Fox-IT, which described how in 2015 one of its researchers spotted a small ‘whitespace’ error in HTTP responses from the ‘beacon’ NanoHTTPD-based web server that can be implanted inside a target network as part of a tool called Cobalt Strike.

Cobalt Strike is a legitimate pen-testing tool used to simulate adversaries in red team testing scenarios. Unfortunately, in recent years it’s also acquired a following among cybercriminals who use it after first breaking its copy protection.

It’s a ready-made platform that gives an adversary (legitimate or otherwise) a foothold through which they can control sideways movement in the network and serve payloads from the comfort of a GUI.

However, the harmless and almost imperceptible whitespace flaw allowed Fox-It to turn this communication into an Intrusion Detection System (IDS) fingerprint which let its analysts see public Cobalt Strike servers.

That remained true until early January, when Cobalt Strike v3.13 finally noticed and fixed an issue which Fox-It believes has been in the software since 2012.

As far as Fox-IT is concerned, this represents a pyrrhic victory for security. Clearly, anything that could, in theory, allow a blue team defender to identify a red team incursion during a pen-testing exercise was going to be removed by Cobalt Strike’s makers.

But assuming cybercriminals implement the update, the fix has also removed the possibility of tracking threat actors using the tool, said to include Carbanak/Fin7, and espionage groups APT29 (Cozy Bear) and China’s APT10 to name only a few.

Indeed, the number of servers featuring the whitespace issue had already declined since the start of 2019. Observes Fox-It:

The change log entry [for v3.13] refers to the removed space being ‘extraneous’, in a literal sense meaning not pertinent or irrelevant. Due to its demonstrated significance as fingerprinting mechanism, this description is contested here.

In total, the company had uncovered 7,718 unique Cobalt Strike team server or NanoHTTPD hosts between January 2015 and February 2019.

Blame game

Is it fair to blame Cobalt Strike for the fact that cybercriminal groups are using it?

Not really. The whole point of pen-testing tools (of which Cobalt Strike is only one) is that the advantages of using them to improve security outweigh any negatives arising from their misuse.

Fox-It recommends that organisations look at the list of whitespace servers it detected to check whether they have been targeted in the past.

Article source: http://feedproxy.google.com/~r/nakedsecurity/~3/8zmWnctUKu8/

Is a Facebookcoin in the works?

Facebook, Signal and Telegram are all planning cryptocurrencies. But why these companies, why now, and will they be successful?

The New York Times published a round-up article looking at the cryptocurrency plans for these three companies, detailing reports that each of them is well along the road to creating its own coins. Facebook’s is the most secretive project.

CEO Mark Zuckerberg has been publicly expressing interest in cryptocurrency since at least January 2018, when he wrote it up in his annual mission statement.

In May that year, he appointed David Marcus, formerly head of the Messenger chat app, to run the company’s blockchain team. Marcus has a history in both fintech and cryptocurrency, having been president at PayPal and spending time on the board of directors at cryptocurrency payments company Coinbase. He resigned that position three months after taking the blockchain lead at Facebook.

Sources told Bloomberg in December that Facebook is reportedly working on a stablecoin, which is a digital currency pegged to a reliable real-world asset like the US dollar. Stablecoins look less like tradable assets that speculators hope will skyrocket in price, and more like proper currencies used to drive everyday transactions. The cryptocurrency would reportedly enable people to transfer money on the WhatsApp messaging system, focusing first on the remittance market in India.

Focusing on India would help the social media giant to foster growth in an emerging market. Beset with branding problems following several privacy missteps in a heavily saturated western market, growth markets like India represent an appealing way for the company to regain ground.

It also seems like a natural evolution for WhatsApp Pay, a service that enables people in India to send each other money directly from their bank accounts via the messaging service.

Introducing cryptocurrency into WhatsApp would accompany a massive shakeup in Facebook’s messaging strategy. The company has announced plans to merge WhatsApp with its Messenger and Instagram apps, making it possible to send messages easily between the services.

The problem facing Facebook is reconciling its aggressively centralised business model, which relies on aggregating and controlling user data, with blockchain technology’s decentralized approach. Zuckerberg is interested in using blockchain for authentication, he said in an interview last month with Harvard Law and computer science professor Jonathan Zittrain:

One of the things that I’ve been thinking about a lot is a use of blockchain that I am potentially interesting [sic] in – although I haven’t figured out a way to make this work out – is around authentication… and basically granting access to your information and to different services. So, basically, replacing the notion of what we have with Facebook Connect with something that’s fully distributed.

However, there is the potential for rogue developers to abuse users beyond the company’s control, he warned, without directly referencing Cambridge Analytica, which did just that.

Telegram

Conversely, Telegram’s plans for a decentralized blockchain system are well understood, being the basis for an original initial coin offering (ICO) that quickly became the most successful ever, raising $1.7 billion in two fundraising rounds last year. The ICO was limited to private accredited investors after the company cancelled the public part of the sale.

Telegram has always planned to release Gram, a cryptocurrency token that could be used to send payments across the encrypted messaging system. It will be part of a shift to a blockchain called the Telegram Open Network (TON), plans for which the company published in a whitepaper. The blockchain infrastructure is now reportedly 90% complete.

Signal

Signal, the other company mentioned in the New York Times’ report, has been planning its MobileCoin since at least December 2011, when it released its whitepaper. The coin, which has its own web site, would operate using the Signal protocol, which WhatsApp also uses. This would make the coin interoperable with WhatsApp and Signal. MobileCoin is advised by Signal founder Moxie Marlinspike, who has significant credibility in the security space having headed security at a Twitter.

These are far from the only companies who have planned their own cryptocurrencies. Reddit also planned its own cryptocurrency token five years ago, according to statements from then-CEO Yishan Wong. The concept of a token, since abandoned, came after investors in a $50m funding round decided to give 10% of the round back to the community.

But what do Telegram, Signal and the Facebook-owned WhatsApp all have in common? They are all highly popular messaging apps looking for new revenue streams. Creating cryptocurrencies of their own would enable them to instantly drive usage to a wide base of messaging users, integrating that usage tightly into a well-understood messaging interface.

Executed well, this would also help to buoy cryptocurrency’s general market standing, moving it from a speculative venture rife with fraud, security, and usability issues to a more mainstream proposition. These efforts are collectively worth watching, and promise to make 2019 a fascinating year for the development of cryptocurrency, a decade after bitcoin first made its debut.

Article source: http://feedproxy.google.com/~r/nakedsecurity/~3/Ez6HP2PUIvw/

TikTok to pay record fine for collecting children’s data

Video streaming app TikTok has agreed to pay a $5.7 million fine for allegedly collecting names, email addresses, pictures and locations of children younger than 13 – all illegal under the US’s Children’s Online Privacy Protection Act (COPPA).

This is the largest settlement ever handed down for violating the nation’s child privacy law, the Federal Trade Commission (FTC) said when it announced the settlement on Thursday.

TikTok, based in Los Angeles, merged with Musical.ly in 2018. The Musical.ly app allowed users to create short videos lip-syncing or dancing to music and to share those videos with other users. Beyond letting users create and share videos, the app also allowed users to interact with other users by commenting on their videos and sending direct messages.

80 million US downloads

TikTok is both massively popular and considered to be addictive. It originally launched in China in 2016, where it was known as Douyin (literally: “vibrating sound”). A year later, it hit the international market with its new name, TikTok.

At least one Chinese doctor specializing in addiction has warned that young people are so hooked on social media approval that they’ve been risking their lives to garner likes with their 15-second Douyin clips, which have featured things like dancing in front of a moving bus or trying to flip a child 180 degrees… and then dropping her.

In April 2018, Douyin launched an anti-addiction system that reminded users to rest after using the app for more than 1.5 hours. When South China Morning Post asked TikTok if it would adopt a similar system, it didn’t reply.

As of June 2018, TikTok said it had 500 million monthly active users worldwide and 150 million daily active users in China. As CNBC reported, it became the world’s most downloaded app on Apple’s App Store in the first half of 2018, with an estimated 104 million downloads: more than that of YouTube, WhatsApp or Instagram for the same period. It’s had 80 million downloads in the US.

Some have raised concerns about privacy on TikTok. In particular, given the app’s popularity with children and teens all over the world, there have been concerns about exposure to sexual predators. That was highlighted last month, when Barnardo’s, a major children’s charity in the UK, found that children as young as eight are being sexually exploited online via social media.

The problem with live streaming video apps such as TikTok is that besides being extremely popular, they’re also very hard to moderate. YouTube, for one, has recently been grappling with a major advertiser backlash against lewd comments being added to videos featuring minors. But an app like TikTok has an added level of moderation difficulty, given that it features real-time comments posted directly to the person streaming: a situation that’s ripe for exploitation.

User accounts public by default

Unfortunately, Musical.ly set the stage for exploitation, given that it collected a lot of information. According to the FTC’s complaint, Musical.ly required users to provide an email address, phone number, username, first and last name, a short biography, and a profile picture. Having public accounts meant that other users could see a child’s profile bio, username, picture, and videos.

Users could change their default setting from public to private, but even so, users’ profile pictures and bios remained public, and other users could still send them direct messages, according to the complaint. There have, in fact, been reports of adults using the app to troll for minors to have sex with.

Beyond that, until October 2016, there was a stalker-friendly feature: the app allowed users to view other users within a 50-mile radius.

What parents might not realize about Musical.ly is that it’s not just a lip-synching/dancing app. It gives users the ability to search specific code words, which leads to hundreds of videos of teenagers stripping or hurting themselves.

Musical.ly allegedly knew it had under-13 users

According to the FTC’s complaint, the operators of Musical.ly knew that they had a “significant” percentage of users younger than 13. They also received thousands of complaints from parents that their children under 13 had created Musical.ly accounts.

The FTC alleges that Musical.ly operators failed to notify parents that they were collecting and using those kids’ personal information, that they never got parental consent before doing so, and that they failed to delete the kids’ information at parents’ request.

Besides the fine, the settlement also requires the app’s operators to comply with COPPA going forward and to remove all videos made by children under the age of 13.

Article source: http://feedproxy.google.com/~r/nakedsecurity/~3/jDj91eMbN1U/

YouTube disables comments on millions of videos of children

On Thursday, YouTube announced on its creator blog that it’s disabling comments on millions of videos featuring minors, in response to reports that creeps are leaving disgustingly sexual comments on videos featuring kids doing things like yoga or gymnastics, or playing games such as Twister.

As content creator Matt Watson had documented a week before, such comments sections had what he called a “wormhole.” Within as few as five clicks, you could find yourself in a “soft-core pedophilia ring” where child oglers leave sexual comments and connect with each other in the comments sections of innocuous videos featuring children, sharing contact information or, sometimes, links to actual child abuse imagery.

The news caused a mob of advertisers to flee, as big brands such as Disney, Fortnite maker Epic Games, GNC and Nestle pulled their ads.

YouTube said on Thursday that over the prior week, the platform had disabled comments on tens of millions of videos that could be “subject to predatory behavior.” Over the coming months, it also plans to suspend comments on videos featuring “young minors” and those featuring older minors that “could be at risk of attracting predatory behavior.”

It’s not shutting down comments on all such videos: YouTube said that a small number of creators will be able to keep comments enabled, though they’ll be required to actively moderate comments and demonstrate a low risk of predatory behavior. YouTube says it’s going to work with such creators directly and hopes that their numbers increase as it works on improving its ability to catch violative comments.

Predator filter will remove 2X more comments

YouTube said that it’s been removing hundreds of millions of violative comments, but it’s been working on an even more effective classifier to specifically sweep up predatory comments. It’s sped up the launch of the classifier, which doesn’t affect video monetization and which is supposed to detect and remove 2X more individual comments.

YouTube said that it’s also removed thousands of inappropriate comments on videos showing minors and has terminated hundreds of viewer accounts for their comments. In addition, it’s reported illegal behavior to the NCMEC so they can work with the proper authorities.

Content creators cry foul

The exodus of big businesses has led some creators to label the incident “Adpocalypse 2.0”.

Adpocalypse 1.0 happened between March-May 2017, when major advertisers such as ATT and Johnson Johnson yanked ads that were appearing on YouTube videos that espoused extremism and hate speech. In response, YouTube abruptly rolled out changes to its automated processes for placing ads across the platform – a move that was considered punitive against creators whose material wasn’t racist or sympathetic to terrorists.

Nearly two years later, content creators again feel like they’re being punished after they’ve done nothing wrong. You’ve got the technology to do more than just blame content creators for this problem, said Twitter user Tay Zonday in response to Google’s announcement:

Others have criticized Watson over his repeated calls for big advertisers to pull out from YouTube and for consumers to boycott them – something he pushed in five videos he uploaded after the initial one.

One critic was Daniel Keem, also known as Keemstar, the host of YouTube channel DramaAlert, who said that a second Adpocalypse wouldn’t stop child abusers:

If advertisers leave YouTube, this isn’t going to stop the pedos in the comments section. This is just going to hurt the livelihood of YouTubers big and small.

Article source: http://feedproxy.google.com/~r/nakedsecurity/~3/gQKwkXMYLeQ/

Ah, this military GPS system looks shoddy but expensive. Shall we try to break it?

Who, Me? Hello, dear readers. We see you’ve come for your weekly dose of Who, Me? to shake off this serious case of the Mondays.

So come with us to 2011, in our weekly column where Reg readers get their worst misdeeds off their chests.

“Joe” was working in the British Army and had been sent a load of trial GPS kit for section level force tracking.

“They came in lovely beige boxes, all secure and sealed with only the push buttons available,” Joe told us.

It was well recognised that this kit was, as Joe put it, “a complete and utter rip-off”.

For instance, it included a “cable tidy” that was made of “high denier cordura [fabric] which cost around £250 per item, when the real cost was less than £50”.

As signal instructors, Joe and his colleague were tasked with sorting through the dozens of boxes, trying to put them in some sort of order and figuring out a bit more about the kit.

“The instructions were pretty poor, but as far as we could tell, each soldier would strap a small box on (about half the size of a VHS box), which held the GPS unit inside it,” he said.

“They’d cable it up to a battery pack which they’d store somewhere else on their person (where, we had no idea).”

The section commander would have all of this, along with another unit – which Joe said was “like the dodgy ’80s tracking devices you see in the movies” with a green screen, circular distance markers and flashing lights to show where troops were.

Swearing and ranting

Blue Monday: Efforts to inspire teamwork with swears back-fires for n00b team manager

READ MORE

“This was about the size and double the thickness of a VHS tape,” said Joe, adding that he had “no idea” why he was using VHS tapes as a unit of measurement. (Perhaps angling to get it included as a Reg Standard?)

Anyway, Joe and his colleague were “pretty pissed off with it all” and were desperate to know what was inside the kit.

“The startup screen on the Command unit showed the old Nokia logo from the 3310 era, and by all accounts could have been something similar,” he said.

So the pair of them decided to do some “light survivability testing”, as he euphemistically branded it.

Of course, the team hadn’t actually been asked to do any such testing, and the units were “really very expensive”.

They were also designed to be worn by soldiers, and presumably get pretty thrashed about.

“Which was why, in the hangar in the tank park, we felt an accurate test would be to drop them from chest height onto the concrete floor,” said Joe.

“They bounced relatively well, a few scrapes, the odd bit of weak corner plastic popping off, but nothing too shabby.”

And so they decided to take it to the next level, and, er, “emulate an angry and frustrated soldier”. By flinging one of the units at the wall, which put a hole in the asbestos-laced barrier, but caused minimal damage to the unit.

By this point, the pair had been sorting through “units and cables and guff” for hours and had both lost the will to live.

“We stood slightly apart, staring in silence at all the units lying on the floor in front of us, when I spotted a sledgehammer leaning against the tank next to me,” said Joe.

“Picking it up by the handle, I looked at my oppo and asked if he thought it’d survive a sledgehammer accidentally falling onto it at speed with force.

“Suffice to say, it didn’t pass that test.”

The pair of them – having finally proved to themselves that, yes, the pricey kit could indeed be destroyed – promptly burst into hysterics and shoved the broken unit into one of the bottom boxes.

“When it was spotted, we said it was like that when we opened the box – shattered, splintered and buggered,” said Joe.

“The kit got sent away, and to this day, I don’t think the very expensive purchase was ever deployed. Another example of MoD waste, which this time we felt we had a particular hand in!”

Have you ever been responsible for wasting taxpayers’ cash and cared not a jot? What about destroying kit for shits ‘n’ giggles? El Reg would love to hear your confessional tales… not to worry, we keep every person’s identity secret. ®

Article source: http://go.theregister.com/feed/www.theregister.co.uk/2019/03/04/who-me/

WannaCry-hero Hutchins’ trial date set, Microsoft readies Google’s Spectre V2 fix for Windows 10, Coinhive axed, and more

Roundup Here’s your weekend rapid-fire roundup of infosec news, ahead of next week’s RSA Conference, beyond what we’ve already covered.

Hutchins’ trial date set: After 18 months in legal limbo in America, Brit malware reverse-engineer Marcus Hutchins, who halted the 2017 Wannacry ransomware outbreak, this week learned he will go before a jury in July.

Hutchins was cuffed in August 2017 in Las Vegas by the FBI, shortly after the global WannaCry infection, and was soon-after formally accused of developing the Kronos banking trojan. He denies any wrongdoing. Since being released on bail, Hutchins has been stuck living on the California coast, and unable to return home to England.

His trial by jury, in a Wisconsin federal district court, is now due to start on July 8. Hutchins has until mid-June to change his plea to guilty, if he so wishes, and have his sentence lessened slightly as a result of avoiding a full-blown expensive trial. His defense costs may hit seven figures, and he is seeking donations to defray costs.

Patch Adobe ColdFusion, Cisco WebEx, Nvidia drivers: Adobe on Friday issued an emergency security update for ColdFusion versions 2018, 2016 and 11 to address a vulnerability (CVE-2019-7816) that can be exploited to execute malicious code on an at-risk installation. This flaw is being targeted right now in the wild by miscreants, we’re told.

“This attack requires the ability to upload executable code to a web-accessible directory, and then execute that code via an HTTP request,” Adobe noted. “Restricting requests to directories where uploaded files are stored will mitigate this attack.”

Also, you probably want to patch Cisco Webex Meetings Desktop App and Cisco Webex Productivity Tools for Windows to address a vulnerability (CVE-2019-1674) that can be exploited to “allow an authenticated, local attacker to execute arbitrary commands as a privileged user.” And Nvidia has emitted a bunch of security fixes to close off arbitrary code execution flaws and escalation-of-privilege blunders, as well as crashes.

Microsoft quietly warms up Google’s Spectre V2 mitigation: In a Windows 10 build 1809 update, KB4482887, issued late this week, Microsoft enabled support for Google’s Retpoline mitigation against Spectre Variant 2 in its kernel, among other bug fixes.

Up until now, Microsoft has relied on processor microcode updates to prevent malware from exploiting Spectre V2 CPU flaws to steal passwords and other secrets from the operating system and other applications. Said microcode patches, simply put, involve repeatedly flushing processor caches to thwart attacks, whereas Retpoline is much more elegant: it changes how software calls subroutines so that it cannot be exploited via Spectre V2.

Crucially, Google’s approach incurs much less of a performance hit than flushing caches all the time, though it requires software be recompiled using the technique. That left Microsoft in a bind: it had to rebuild, or patch on the fly, its operating system to make use of Google’s breakthrough, and that still left third-party closed-source kernel-mode drivers vulnerable to exploitation. Until now, Retpoline has remained disabled by default in Windows 10 for the vast majority of users, who rely instead on microcode patches, though it has been available to some Insider testers.

Now, with this update, the latest edition of Windows 10 can use fast Retpoline where possible, and fall back to slow cache flushing when it can’t due to vulnerable third-party drivers and so on. Retpoline has about a two-per-cent overhead, whereas the microcode approach is many times that, depending on the workload.

Microsoft, refreshingly, goes into much more technical detail on the changes here. Essentially, if you’re running Windows 10 build 1809, aka the big October 2018 upgrade, look out for this update and install it once you’re happy with it, so as to eventually benefit from Retpoline’s performance boost. The changes are also expected to be baked into Windows 10 19H1, due out this Spring.

It also sounds as though Microsoft will gradually enable Retpoline for users, taking it nice and slow rather than breaking tens or hundreds of millions of installations at once, because it involves fundamentally changing the way its operating system branches to subroutines. “Over the coming months, we will enable Retpoline as part of phased rollout via cloud configuration,” the biz explained in its tech notes. “Due to the complexity of the implementation and changes involved, we are only enabling Retpoline performance benefits for Windows 10, version 1809 and later releases.”

D’oh Jones! News database exposed online: A copy of Dow Jones’ Watchlist – a paid-for database of news articles and other public sources on politicians, terrorists, criminals, their friends and families, and other such interesting folks – was accidentally left facing the internet. The poorly secured AWS Elasticsearch data silo, containing 2,418,862 records, has since been hidden from view.

“This data is entirely derived from publicly available sources,” a Dow Jones spokesperson told Bob Diachenko, who discovered the cockup and flagged it up this week. “At this time our review suggests this resulted from an authorized third party’s misconfiguration of an AWS server, and the data is no longer available.”

Crypto-bête-noire Coinhive to shut down: Coinhive, makers of JavaScript that secretly mines Monero crypto-currency that’s been embedded in countless pages of hacked and non-hacked websites and web apps, is said to be pulling the plug on March 8.

“It has been a blast working on this project over the past 18 months, but to be completely honest, it isn’t economically viable anymore,” its operators wrote this week.

“The drop in hash rate (over 50%) after the last Monero hard fork hit us hard. So did the ‘crash’ of the crypto-currency market with the value of XMR depreciating over 85% within a year. This and the announced hard fork and algorithm update of the Monero network on March 9 has lead us to the conclusion that we need to discontinue Coinhive.

“Thus, mining will not be operable anymore after March 8, 2019. Your dashboards will still be accessible until April 30, 2019 so you will be able to initiate your payouts if your balance is above the minimum payout threshold.”

Huawei bean-counter extradition hearing green-light: Canadian authorities have decided to put America’s extradition request for Huawei CFO Meng Wanzhou before a judge. The hearing is set to take place on March 6.

DEF CON call for papers: This year’s DEF CON hacking conference is now accepting proposals for talks, and has offered to cover hotel bills for up to three nights.

DDoS-for-hire bloke ‘fesses up: Sergiy P. Usatyuk, 20, of Orland Park, Illinois, in the US, pleaded guilty this week to conspiracy to cause damage to internet-connected computers by launching distributed-denial-of-service attacks against victim’s internet connections and websites in exchange for money. Usatyuk and a co-conspirator banked more than $550,000 from knocking netizens and organizations offline, according to prosecutors.

DNSSEC push renewed: DNS overlord ICANN has urged net admins to deploy DNSSEC technology to protect websites from being hijacked by miscreants, following a spate of domain takeovers. These hijackings are typically the result of crooks breaking into weakly secured domain registrar user accounts, rather than exploiting underlying protocols and systems.

Pubs, hotels’ payment systems hacked: If you’ve paid for anything at these bars, restaurants, and hotels in America between January 3 and 24 this year, using a debit or credit card, then the details – the cardholder’s name, card number, card expiration date, and CVV – were probably snaffled by malware on the payment systems, and siphoned off to fraudsters.

The affected businesses are spread out over Arizona, Minnesota, Louisiana, Iowa, Missouri, North and South Dakota, Texas, Wisconsin, Tennessee, Oregon, California, Colorado, and Ohio. ®

Article source: http://go.theregister.com/feed/www.theregister.co.uk/2019/03/02/security_roundup/

WannaCry-hero Hutchins’ trial date set, Microsoft readies Google’s Spectre V2 fix for Windows 10, Coinhive axed, and more

Roundup Here’s your weekend rapid-fire roundup of infosec news, ahead of next week’s RSA Conference, beyond what we’ve already covered.

Hutchins’ trial date set: After 18 months in legal limbo in America, Brit malware reverse-engineer Marcus Hutchins, who halted the 2017 Wannacry ransomware outbreak, this week learned he will go before a jury in July.

Hutchins was cuffed in August 2017 in Las Vegas by the FBI, shortly after the global WannaCry infection, and was soon-after formally accused of developing the Kronos banking trojan. He denies any wrongdoing. Since being released on bail, Hutchins has been stuck living on the California coast, and unable to return home to England.

His trial by jury, in a Wisconsin federal district court, is now due to start on July 8. Hutchins has until mid-June to change his plea to guilty, if he so wishes, and have his sentence lessened slightly as a result of avoiding a full-blown expensive trial. His defense costs may hit seven figures, and he is seeking donations to defray costs.

Patch Adobe ColdFusion, Cisco WebEx, Nvidia drivers: Adobe on Friday issued an emergency security update for ColdFusion versions 2018, 2016 and 11 to address a vulnerability (CVE-2019-7816) that can be exploited to execute malicious code on an at-risk installation. This flaw is being targeted right now in the wild by miscreants, we’re told.

“This attack requires the ability to upload executable code to a web-accessible directory, and then execute that code via an HTTP request,” Adobe noted. “Restricting requests to directories where uploaded files are stored will mitigate this attack.”

Also, you probably want to patch Cisco Webex Meetings Desktop App and Cisco Webex Productivity Tools for Windows to address a vulnerability (CVE-2019-1674) that can be exploited to “allow an authenticated, local attacker to execute arbitrary commands as a privileged user.” And Nvidia has emitted a bunch of security fixes to close off arbitrary code execution flaws and escalation-of-privilege blunders, as well as crashes.

Microsoft quietly warms up Google’s Spectre V2 mitigation: In a Windows 10 build 1809 update, KB4482887, issued late this week, Microsoft enabled support for Google’s Retpoline mitigation against Spectre Variant 2 in its kernel, among other bug fixes.

Up until now, Microsoft has relied on processor microcode updates to prevent malware from exploiting Spectre V2 CPU flaws to steal passwords and other secrets from the operating system and other applications. Said microcode patches, simply put, involve repeatedly flushing processor caches to thwart attacks, whereas Retpoline is much more elegant: it changes how software calls subroutines so that it cannot be exploited via Spectre V2.

Crucially, Google’s approach incurs much less of a performance hit than flushing caches all the time, though it requires software be recompiled using the technique. That left Microsoft in a bind: it had to rebuild, or patch on the fly, its operating system to make use of Google’s breakthrough, and that still left third-party closed-source kernel-mode drivers vulnerable to exploitation. Until now, Retpoline has remained disabled by default in Windows 10 for the vast majority of users, who rely instead on microcode patches, though it has been available to some Insider testers.

Now, with this update, the latest edition of Windows 10 can use fast Retpoline where possible, and fall back to slow cache flushing when it can’t due to vulnerable third-party drivers and so on. Retpoline has about a two-per-cent overhead, whereas the microcode approach is many times that, depending on the workload.

Microsoft, refreshingly, goes into much more technical detail on the changes here. Essentially, if you’re running Windows 10 build 1809, aka the big October 2018 upgrade, look out for this update and install it once you’re happy with it, so as to eventually benefit from Retpoline’s performance boost. The changes are also expected to be baked into Windows 10 19H1, due out this Spring.

It also sounds as though Microsoft will gradually enable Retpoline for users, taking it nice and slow rather than breaking tens or hundreds of millions of installations at once, because it involves fundamentally changing the way its operating system branches to subroutines. “Over the coming months, we will enable Retpoline as part of phased rollout via cloud configuration,” the biz explained in its tech notes. “Due to the complexity of the implementation and changes involved, we are only enabling Retpoline performance benefits for Windows 10, version 1809 and later releases.”

D’oh Jones! News database exposed online: A copy of Dow Jones’ Watchlist – a paid-for database of news articles and other public sources on politicians, terrorists, criminals, their friends and families, and other such interesting folks – was accidentally left facing the internet. The poorly secured AWS Elasticsearch data silo, containing 2,418,862 records, has since been hidden from view.

“This data is entirely derived from publicly available sources,” a Dow Jones spokesperson told Bob Diachenko, who discovered the cockup and flagged it up this week. “At this time our review suggests this resulted from an authorized third party’s misconfiguration of an AWS server, and the data is no longer available.”

Crypto-bête-noire Coinhive to shut down: Coinhive, makers of JavaScript that secretly mines Monero crypto-currency that’s been embedded in countless pages of hacked and non-hacked websites and web apps, is said to be pulling the plug on March 8.

“It has been a blast working on this project over the past 18 months, but to be completely honest, it isn’t economically viable anymore,” its operators wrote this week.

“The drop in hash rate (over 50%) after the last Monero hard fork hit us hard. So did the ‘crash’ of the crypto-currency market with the value of XMR depreciating over 85% within a year. This and the announced hard fork and algorithm update of the Monero network on March 9 has lead us to the conclusion that we need to discontinue Coinhive.

“Thus, mining will not be operable anymore after March 8, 2019. Your dashboards will still be accessible until April 30, 2019 so you will be able to initiate your payouts if your balance is above the minimum payout threshold.”

Huawei bean-counter extradition hearing green-light: Canadian authorities have decided to put America’s extradition request for Huawei CFO Meng Wanzhou before a judge. The hearing is set to take place on March 6.

DEF CON call for papers: This year’s DEF CON hacking conference is now accepting proposals for talks, and has offered to cover hotel bills for up to three nights.

DDoS-for-hire bloke ‘fesses up: Sergiy P. Usatyuk, 20, of Orland Park, Illinois, in the US, pleaded guilty this week to conspiracy to cause damage to internet-connected computers by launching distributed-denial-of-service attacks against victim’s internet connections and websites in exchange for money. Usatyuk and a co-conspirator banked more than $550,000 from knocking netizens and organizations offline, according to prosecutors.

DNSSEC push renewed: DNS overlord ICANN has urged net admins to deploy DNSSEC technology to protect websites from being hijacked by miscreants, following a spate of domain takeovers. These hijackings are typically the result of crooks breaking into weakly secured domain registrar user accounts, rather than exploiting underlying protocols and systems.

Pubs, hotels’ payment systems hacked: If you’ve paid for anything at these bars, restaurants, and hotels in America between January 3 and 24 this year, using a debit or credit card, then the details – the cardholder’s name, card number, card expiration date, and CVV – were probably snaffled by malware on the payment systems, and siphoned off to fraudsters.

The affected businesses are spread out over Arizona, Minnesota, Louisiana, Iowa, Missouri, North and South Dakota, Texas, Wisconsin, Tennessee, Oregon, California, Colorado, and Ohio. ®

Article source: http://go.theregister.com/feed/www.theregister.co.uk/2019/03/02/security_roundup/

After last year’s sexism shambles, 2019’s RSA infosec bash has upped its inclusivity game

RSA As San Francisco gets ready for its annual RSA gabfest Conference, taking place next week, organisers appear to have got the message over inclusivity following last year’s fiasco.

When the 2018 event was announced, many in the infosec industry were shocked that, despite the wealth of talent across all genders in the sector, precisely one non-male keynote speaker had been booked – Monica Lewinsky. The matter wasn’t helped by RSA’s initial response that computer security was male-dominated, and they just couldn’t find good women speakers.

This was embarrassingly proved to be garbage when a group of volunteers, operating on a shoestring budget and working in their spare time, managed to get the OURSA conference up and running, with 14 senior women in the cybersecurity industry giving a range of talks. This practical demonstration seems to have shamed RSA into action, and now there’s a much more diverse lineup of speakers for this year’s shindig.

“They have made a lot of progress in gender parity this year,” Melanie Ensign, Uber’s head of security and privacy communications, and one of the OURSA conference organizers, told The Register.

“What’s missing is the overall acknowledgement of the environment and culture of the conference. When you think about culture, it doesn’t matter how many women are on the show’s stages if I’m getting harassed on the show floor. I know they can do better, they have the resources and means to do it.”

Someone else’s problem

One senior woman security executive, who wished to remain anonymous, reckoned this isn’t entirely RSA’s fault. The conference organisers auction off many of the keynote spots to the highest bidder, and it’s up to the paying companies to decide who they send – and in the past, they’ve put mostly blokes on the stage.

“Companies think of it as RSA’s problem that there aren’t diverse speakers,” she told The Reg. “Everyone else is waiting for them to solve the problem because they don’t want to give up time on stage. That’s one of the reasons why RSA doesn’t have more female speakers.”

As she and others have pointed out, RSA isn’t really a security conference as such, but a sales bonanza where the security industry tries to shift kit. While the exhibition floor has been noticeably lacking in booth beefcakes and babes, there hasn’t been a RSA yet that this hack hasn’t heard tales of women being harassed on the show floor.

Microsoft dancers at GDC

Microsoft’s equality and diversity: Skimpy schoolgirls dancing for nerds at an Xbox party

READ MORE

This isn’t just a problem for RSA, but one that bedevils many tech events. In recent years the adoption of strict codes of conduct have helped matters, though there’s still a lot of work to be done.

“RSA was a focal point for us last year, but we were also trying to demonstrate the value of diverse content and people to the entire industry,” said Alex Stamos, another key player in OURSA who is currently recovering from his stint as Facebook’s CSO as an adjunct professor at Stanford.

“It’s great that RSA is fixing things, but much of the infosec circuit seems to be moving backwards.” ®

Article source: http://go.theregister.com/feed/www.theregister.co.uk/2019/03/01/rsa_speaker_lineup_inclusivity/

Did you hear the one about Cisco routers using strcpy insecurely for login authentication? Makes you go AAAAA-AAAAAAArrg *segfault*

Cisco has patched three of its RV-series routers after Pen Test Partners (PTP) found them using hoary old C function strcpy insecurely in login authentication function. The programming blunder can be exploited to potentially hijack the devices.

PTP looked at how the routers’ web-based control panel handled login attempts by users, and found that it was alarmingly easy to trigger a buffer overflow by simply supplying a long string of characters as the password, something which Cisco admitted “could allow an unauthenticated, remote attacker to execute arbitrary code on an affected device”.

Lobbing in a password of 447 characters, such as ‘A’, followed by four characters, would allow the hijacker to control a subroutine return address on the web app’s stack using the values of those final characters. That means the hacker could force the device’s 32-bit Arm-based processor to jump to malicious code stashed in the login request.

An attacker has to be able to reach the web interface – either on the local network, or via the remote administration feature that is admittedly disabled by default.

The three routers affected – the RV110W, RV130W and RV215W – run “some form of embedded Linux” instead of Cisco OS, according to PTP’s definitely-not-pseudonymous blogger Dave Null. The network equipment provides physical Ethernet ports as well as Wi-Fi connectivity, firewall protections, and VPN tunneling.

The discovery of the flaw was credited to Yu Zhang and Haoliang Lu at the GeekPwn conference, and T. Shiomitsu of Pen Test Partners, who worked and informed Cisco separately.

Insecurity a-bounds

When following the RV130W’s login process at the binary executable level, PTP found that the router was placing the user-inputted password string into a memory buffer, ready for authentication against the saved password, using strcpy with no bounds checks. That means if the password is larger than the buffer size, it will blow past the end of the buffer and overwrite other data and return addresses on the stack.

As El Reg reported years ago when a similarly worrying use of strcpy emerged in glibc: “strcpy is dangerous and an obvious target in an audit because it blindly copies the entire contents of a zero-terminated buffer into another memory buffer without checking the size of the target buffer.”

Null from PTP elaborated on Thursday: “If someone else has control over the source string, you are giving an external entity the capability to overwrite the bounds of the memory that you allocated – which might mean they can overwrite something important with something bad. In most exploitable cases, this will mean overwriting a saved return pointer on the stack and redirecting the execution flow of the process.”

He cheerily added: “Oh yeah, also, no PIE/ASLR in the binary,” meaning it’s pretty easy to exploit. Tutorials on how to leverage buffer overflows on embedded Arm-based gadgets can be found here, for instance.

The security patches for the hardware are free to download, and were released earlier this week; you don’t need a support contract. Cisco customers should check their routers are running the latest firmware versions, to ensure they are not vulnerable, as follows:

  • RV110W Wireless-N VPN Firewall: 1.2.2.1
  • RV130W Wireless-N Multifunction VPN Router: 1.0.3.45
  • RV215W Wireless-N VPN Router: 1.3.1.1

A decade ago Microsoft banned the use of a superficially similar function, memcopy, from its code. PTP’s Null suggested latterday C authors might want to switch to strlcpy instead, “a nonstandard function which takes a third length argument, and always null terminates”. ®

Article source: http://go.theregister.com/feed/www.theregister.co.uk/2019/03/01/cisco_cve_2019_1663_strcpy_login_authentication/

When the bits hit the FAN: US military accused of knackering Russian trolls, news org’s IT gear amid midterm elections

A Russian new service is claiming that US attacks on it and an organisation accused of state-sponsored trolling has left storage systems damaged and international servers wiped after multiple malware attacks.

The Russian Federal News Agency (FAN) alleged earlier this week that US Cyber Command conducted an online attack against the self-described news organization in conjunction with Cyber Command’s reported offensive operation against the Internet Research Agency (IRA), a Kremlin-linked Russian organization based in St Petersburg that US officials blame for spreading misinformation through social media to sow discord and interfere with elections.

The report lends support to claims that the US military conducted offensive cyber operations in Russia last year to prevent interference with the 2018 midterm elections.

FAN, according to the US government, is a part of Project Lakhta, the Russian-funded foreign influence operation. The publisher is said to have served as a conduit for concealing Project Lakhta activities and as such is currently subject to Treasury Department sanctions. The IRA is also said to be part of Project Lakhta. FAN nonetheless maintains it has nothing to do with the IRA and does not interfere with elections or engage in illegal activity.

Anatomy of a hack

Two weeks ago, General Paul Nakasone, head of US Cyber Command, hinted at Russia-focused operations in testimony to the US Senate Armed Services Committee. The military cyber group, he said, “undertook an initiative known as the Russia Small Group to protect the elections from foreign interference and influence.”

On Tuesday, The Washington Post reported that US Cyber Command meddled with the Internet Research Agency’s ability to access the internet during last year’s midterms. No details about the nature of the disruption were disclosed, however.

FAN’s unconfirmed account may help fill in some of the blanks, though bear in mind the organization is not considered a reliable news source in the West. Here’s what the network claimed:

On November 5, 2018 at about 22:00 hours Moscow time, the RAID controller on the internal office of the FAN was destroyed and two out of four hard drives were disabled. The hard drives were also formatted on servers leased in Sweden and Estonia that were used to store data from the USA Really portal.

FAN claims that US Cyber Command sent an employee a malicious email attachment that installed malware on a Windows machine, but that network security measures prevented the intrusion from doing harm beyond that single machine. But this was just the start, FAN claims.

ira

Bloke gets six months for fixing up Russia’s US election trolls with bank accounts, fake identities

READ MORE

An Apple iPhone 7 Plus, the organization says, is what allowed the attackers to access FAN’s local network. According to the group, an employee’s iPhone automatically launched iTunes when connected to a USB cable, prompting synchronization and Windows updates on the host PC, which apparently allowed the takeover of the connected computer.

The firm contends that the intrusion came from IP addresses associated with American companies, including Amazon – which, remember, runs a cloud service, AWS. And it says that it has revised its corporate security policy to prohibit connecting iPhones to computers.

Despite this, FAN characterized the US cyber attack as a failure due to the lack of trumpeting about the incident from US authorities.

The US Defense Department declined to comment. “We do not discuss classified cyberspace operations due to classification and operations security,” a DoD spokesperson told The Register via email. ®

Article source: http://go.theregister.com/feed/www.theregister.co.uk/2019/03/01/russian_ira_hack/