Information Technology

Facebook announced on Monday that it shut down a “troll farm” allegedly run by the government of Nicaragua and the Sandinista National Liberation Front (FSLN) party. Nicaraguan President Daniel Ortega is vying for a fourth consecutive term in an election on Sunday.The company — which recently adopted the new name of “Meta” — made the announcement in its “October 2021 Coordinated Inauthentic Behavior Report.”Facebook says that in October, it removed 937 Facebook accounts, 363 Instagram accounts, 140 pages and 24 groups connected to the campaign. All of the accounts, pages and groups were allegedly connected to people in Nicaragua. They called it “one of the most cross-government troll operations we’ve disrupted to date,” and said multiple state entities were involved.”This operation targeted domestic audiences in that country and was linked to the government of Nicaragua and the Sandinista National Liberation Front (FSLN) party. We found one portion of this network through our internal investigation into suspected coordinated inauthentic behavior in the region, and another portion — as a result of reviewing public reporting about some of this activity,” Facebook said. “Our teams continue to focus on finding and removing deceptive campaigns around the world — whether they are foreign or domestic. We know that influence operations will keep evolving in response to our enforcement, and new deceptive behaviors will emerge. We will continue to refine our enforcement and share our findings publicly. We are making progress rooting out this abuse, but as we’ve said before — it’s an ongoing effort and we’re committed to continually improving to stay ahead.”Facebook defines “coordinated inauthentic behavior” as attempts to “manipulate public debate” through fake accounts. 

The company claims it is working to stop campaigns run by governments and non-governmental groups, adding that it watches for “efforts to re-establish a presence on Facebook by networks we previously removed.” They use manual and automated tools to detect the campaigns. “The use of government employees and infrastructure to run large-scale, cross-platform troll operations is an especially troubling trend: this year alone, we have taken down government-linked CIB networks in Ethiopia, Uganda, Sudan, Thailand and Azerbaijan,” Facebook said. Facebook did not respond to questions about why they identified and spotlighted this specific campaign considering the many similar campaigns run by government actors in dozens of countries. The Nicaraguan government and FSLN also did not respond to requests for comment. Facebook’s IO Threat Intelligence Team, led by Luis Fernando Alonso and Ben Nimmo, claimed in a report that the campaigns began in April 2018. “It was primarily operated by employees of the Nicaraguan Institute of Telecommunications and the Post (TELCOR), working from the headquarters of the postal service in Managua.”Additional smaller clusters of fake accounts were run from other government institutions, including the Supreme Court and the Nicaraguan Social Security Institute. This campaign was cross-platform as well as cross-government. It ran a complex network of media brands across Facebook, Tiktok, Instagram, Twitter, YouTube, Blogspot and Telegram, as well as websites tied to these news entities. They posted positive content about the government and negative commentary about the opposition, using hundreds of fake accounts to promote these posts.”The Facebook team noted that this campaign was “distinct from financially-motivated clickbait content farms which don’t necessarily rely on fake accounts, but rather use Pages and Groups to post clickbait to drive people to off-platform websites and other channels to monetize.”The campaign purportedly began in response to student-led nationwide protests and allegedly involved spreading information that “focused on discrediting the protesters, dissemination of false information and mass reporting of people opposing the government.”According to Facebook, the effort switched from criticizing protesters to promoting the government’s work in the country. The pages involved in the campaign had about 585,000 followers and nearly 74,500 joined the groups involved. About 125,000 accounts followed the Instagram accounts involved, according to Facebook. The campaign also involved coordinated attempts to report the posts and pages of government critics to Facebook as a way to have them taken down. “These included activists, independent media outlets and regular members of the public who had criticized government policies. Our review of these reports suggests that the great majority were rejected,” Facebook said. “In at least one case, the network tried to get a series of posts that exposed its activity taken down, including photos of an apparent troll facility inside the TELCOR building in Managua. This attempt, too, failed. Despite being mostly unsuccessful, this tactic highlights how the organization sought to control the information environment of everyday Nicaraguan citizens. Although the operators posed as regular citizens of Nicaragua, our investigation found that much of the activity was operated from government-linked entities in Managua, including TELCOR.”Facebook continues to face withering backlash for its failure — or in some cases complicity — in relation to government campaigns using the platform for nefarious purposes. Former Facebook employee Frances Haugen leaked thousands of documents showing that Facebook has markedly different moderation policies in different parts of the world and that CEO Mark Zuckerberg specifically made decisions based on politics as opposed to the best interest of users and the public, like censoring anti-government posts in Vietnam and opposing publishing Spanish-language voting information in the US. More

CrowdStrike announced on Monday that it acquired SaaS-based cybersecurity service SecureCircle in an all-cash deal expected to close during the company’s fiscal fourth quarter.Terms of the deal were not disclosed, but CrowdStrike said the acquisition will allow them to “extend Zero Trust security to data on the endpoint.”George Kurtz, co-founder and chief executive officer of CrowdStrike, said data loss prevention has suffered from a lack of innovation and he noted that legacy tools have failed to live up to the promise of preventing breaches. “At the same time, the endpoint has become the focal point for how data is accessed, used, shared and stored,” Kurtz said. “CrowdStrike will be setting a new standard for endpoint-based data protection by connecting Zero Trust enforcement to the device, the user identity and, with this acquisition, the data users are accessing and using.”CrowdStrike explained in a statement that SecureCircle’s technology will help them “modernize data protection and enable customers to enforce Zero Trust at the device level, the identity level, and at the data level.”The company specifically cited the effect SecureCircle’s tools will have on CrowdStrike’s Falcon agent when securing the endpoint. SecureCircle’s technology helps customers enforce encryption on data in transit, at rest and in use, and CrowdStrike called data loss prevention a “failed technology” as companies continue to deal with data breaches on a daily basis. 

The combination of tools, according to the two companies, will allow users to control usage policies for data and access rules. “We are excited to join the CrowdStrike family, and integrate SecureCircle’s revolutionary data protection solutions with the industry leader in cloud-delivered endpoint protection,” said Jeff Capone, chief executive officer at SecureCircle. “The endpoint in today’s enterprise is everything, and coupling our cloud-native approach to protecting sensitive data with CrowdStrike’s industry leading Zero Trust endpoint security will enable customers to enforce Zero Trust on the endpoint across all levels.” More

The Cring ransomware group continues to make a name for itself through attacks on aging ColdFusion servers and VPNs after emerging earlier this year. 

Experts like Digital Shadows Sean Nikkel told ZDNet that what makes Cring interesting is that so far, they appear to specialize in using older vulnerabilities in their attacks. “In a previous incident, Cring operators exploited a two-year-old FortiGate VPN vulnerability to target end-of-life Microsoft and Adobe applications. This should be a wake-up call for system owners everywhere who are using end-of-life or otherwise unsupported systems that are exposed to the internet at large,” Nikkel said. “While Cring has operators that have used Mimikatz on systems to gain credentials, there’s also evidence of native Windows process usage, which potentially blends in with otherwise legitimate activity. This can often make it more tricky for network hunters and defenders to see anything malicious until it’s too late. This and previous attacks also showcase the continued adoption and use of Cobalt Strike beacons by various threat actors, which often make the post-exploit phase easier for attackers to manage.”Sophos released a report in September highlighting one specific incident where Cring operators exploited a vulnerability in an 11-year-old installation of Adobe ColdFusion 9 to take control of a ColdFusion server remotely. Sophos was able to tie the group using the Cring ransomware to hackers in Belarus and Ukraine that used automated tools to break into the servers of an unnamed company in the services sector. The hackers used their automated tools to browse 9,000 pathways into the company’s systems in 75 seconds. Three minutes later, they were able to exploit a vulnerability in the outdated Adobe program that allowed them to get their hands on files from servers that weren’t supposed to be publicly available. They grabbed a file called “password properties,” and wrote garbled code on top of their “footprints” to cover their tracks. Then, they waited two and a half days, came back into the company’s network, gave themselves Admin privileges and posted a sardonic ransom note. 

The hackers were also able to get access to timesheets and accounting data for payroll before breaching the internet-facing server in minutes and executing the ransomware 79 hours later.Andrew Brandt, principal researcher at Sophos, said the Cring ransomware isn’t new, but it’s uncommon. “In the incident we researched, the target was a services company, and all it took to break in was one internet-facing machine running old, out-of-date and unpatched software. The surprising thing is that this server was in active daily use. Often the most vulnerable devices are inactive or ghost machines, either forgotten about or overlooked when it comes to patching and upgrades,” Brandt said. “But, regardless of what the status is — in use or inactive — unpatched internet-facing servers or other devices are prime targets for cyberattackers scanning a company’s attack surface for vulnerable entry points. This is a stark reminder that IT administrators benefit from having an accurate inventory of all their connected assets and cannot leave out-of-date critical business systems facing the public internet. If organizations have these devices anywhere on their network, they can be sure that cyberattackers will be attracted to them. Don’t make life easy for cybercriminals.”The attack identified by Sophos found that the hackers scanned the victim’s website with automated tools and gained easy access once they found the unpatched ColdFusion on a server. Sophos researchers noted that the Cring operators “used fairly sophisticated techniques to conceal their files, inject code into memory, and cover their tracks by over-writing files with garbled data or deleting logs and other artifacts that threat hunters could use in an investigation.” After getting around security features, the hackers left a note saying, “ready to leak in case we can not make a good deal.”Pavel Kuznetsov, deputy managing director for cybersecurity technologies for Positive Technologies, told ZDNet that Cring operators are routinely interested in conducting sufficiently deep reconnaissance inside the network before direct infection by their ransomware. “Among the targets are often the infrastructures of industrial organizations. Moreover, ICS segments are selected for infection by the ransomware, obviously with the aim of endangering the associated processes (production, etc.),” Kuznetsov said. Positive Technologies head of malware detection Alexey Vishnyakov added that the group gets its primary consolidation through the exploitation of 1-day vulnerabilities in services at the perimeter of the organization like web servers, VPN solutions and more, either through buying access from intermediaries on shadow forums or other methods. “The group uses Mimikatz to move inside an organization. It uses the Cobalt Strike pentesting tool to secure it within the network to the hosts. After taking over the network, it downloads and distributes the ransomware,” Vishnyakov said. Vishnyakov echoed Kuznetsov’s analysis that Cring was focused on attacking industrial companies, hoping to force suspensions of production processes and financial losses as a way to push victims into paying ransoms. “It is far from the first and won’t be the last criminal group that acts according to the scheme of compromising an unpatched vulnerability and encrypting data,” Vishnyakov said. “Particularly dangerous is a series of successful penetrations and production infections. Risks include not only blackmail and financial consequences — these attacks could also possibly lead to accidents and death. More

ZDNet Recommends

Ransomware decryptors for the BlackByte, Atom Silo, LockFile and Babuk strains were released over the last two weeks, highlighting some amount of progress in the fight against a few of the smaller ransomware gangs.Last week, security company Avast released three decryptors, including ones for those affected by the AtomSilo, LockFile and Babuk ransomware. Cybersecurity firm Trustwave released a decryptor for the BlackByte ransomware two weeks ago.  Allan Liska, a ransomware expert with the Recorded Future security company, told ZDNet that while it often feels as though security teams are losing the fight against ransomware, there is progress being seen. “Since August, by my count, we have seen decryptors for BlackMatter, REvil, AtomSilo, Babuk, LockFile, BlackByte, Prometheus and Ragnarok (I’m probably missing some others),” Liska said.When asked why there was a recent wave of decryptors being released, Liska attributed it to a number of factors.”Security researchers at companies like Emsisoft are getting better at finding flaws in ransomware and writing decryptors. And, communication between security companies on ransomware is increasing, so we are sharing information privately that helps victims,” Liska said.”We can’t discount the impact that 10 law enforcement actions against ransomware groups is also having. These actions, like the ones recently, raise the cost of ransomware operations, and many third and fourth-tier ransomware groups are deciding it is no longer worth the risk. So, they ‘retire’ and release their keys, making it easier to create decryptors.”

BreachQuest CTO Jake Williams noted that each of the most recent ransomware decryptors released was enabled by operational security or programming mistakes made by the threat actors. Security teams, he added, had little to do with this recent wave of decryptors other than the possibility that they were getting better at operationalizing available data.”The LockFile/AtomSilo decryptor targets weaknesses in the implementation of the cryptographic algorithm used to encrypt the files. The same is true for BlackByte. In the case of Babuk, the decryptor was enabled by a source code dump in September. It’s worth noting that any encryptions performed by Babuk after the source code dump probably can’t be decrypted by the tool. This is because the master key has been changed after leaking in the dump,” Williams explained.When asked which recent decryptor would be the most consequential, Williams said that without a doubt, it would be Babuk. “This wasn’t enabled by any cryptographic weaknesses and instead required a leak. The fact that the ransomware source code leaked at all is likely driving anxiety in the ransomware operator community, which in recent months has also seen the leak of the Conti ransomware affiliate handbook and successful law enforcement action against REvil,” Williams told ZDNet. Like Digital Shadows’ Ivan Righi, other experts said malware analysts are improving and capitalizing on mistakes or weaknesses in threat actors’ encryption processes. Ransomware has been a big focus for many security teams in 2020 and 2021, and the more resources that are invested into fighting against ransomware, the smaller the room for mistakes is from a cybercriminal’s perspective, Righi said. Over the last few years, the wealth brought in by certain ransomware gangs has attracted multiple threat actors, some of whom are not as sophisticated as others.”As the number of ransomware variants continues to pile up, it is no surprise that we will begin to find weaknesses in some of these ransomware variants, which may allow for decryption keys to be extracted,” Righi said. Of all the decryptors released over the past few months, the universal decryptor for victims of the Kaseya VSA supply-chain attack stood out most to Righi.Released in September by Bitdefender, the universal decryptor only works for REvil/Sodinokibi victims infected before July 13, 2021. Hundreds of victims were helped with the decryptor after the group went dark yet again last month. It was later revealed that law enforcement officials from multiple countries were involved in disrupting the REvil ransomware gang.But Righi noted that just because a decryptor is released, that doesn’t mean a ransomware gang is necessarily finished. “The release of a decryptor for a ransomware variant does not mean the end of that ransomware group. DarkSide had a decryptor released in January 2021, but the group simply improved its tools and continued attacks until May 2021, when the Colonial Pipeline attack occurred,” Righi said. “However, the release of decryptors may damage a group’s reputation and ability to attract new affiliates.” More

In Forrester’s Predictions blog last year, our security team talked about 2021 as “the transition to a new normal.” It hasn’t quite worked out that way — as the Delta variant spread and lockdowns reappeared, employees who had flirted with heading back to the office found themselves joining remote meetings from home just like before. As we look forward to 2022, a theme emerges: securing relationships. Uncertainty has accelerated reliance on each other, and gaps in third-party relationships, collaboration, and trust will have outsized impacts on firms’ relationships with their colleagues, partners, and suppliers.  For cybersecurity, here’s what we expect to see in 2022: Sixty percent of security incidents will result from issues with third parties. In 2020, 27.8% of organizations reported 20 or more supply chain disruptions, and executives have uttered the phrase “supply chain” over 3,000 times on S&P 500 earnings calls, compared to 2,100 times all of last year. A quick look at Google Trends reveals that searches for “supply chain” have peaked just in the last couple of weeks. With cyberattacks targeting smaller vendors and suppliers, third-party incidents will increase and SolarWinds-style headlines will plague firms that don’t invest in the risk management trifecta: people, process, and technology. Security brain drain sets in as one in 10 experienced security pros exit the industry. Two million women have left the US labor force during the pandemic according to data from the US Labor Department, roughly twice as many as men. That’s a sobering trend for an industry like cybersecurity which is already struggling with diversity, equity, and inclusion as well as burnout. Data in a 2021 study from VMware shows that 51% of cybersecurity professionals experienced extreme stress or burnout over the past twelve months. CISOs must tackle the problems of burnout and team culture while using succession planning to build a pipeline of future security leaders. At least one security vendor collapses in an Enron-Theranos-esque scandal. Eighteen cybersecurity vendors reached unicorn status in the first half of 2021, compared to only six the entire previous year. For context, a year before Cisco acquired Duo Security, Duo’s most recent valuation had put it just above unicorn status at $1.17 billion. With the explosion of investment activity in cybersecurity, we expect “accounting irregularities” will bring at least one cybersecurity vendor down in 2022. The fallout creates risks for CISOs. To reduce these risks, when working with early-stage security startups, consider adding a second vendor for redundancy, and take a cautious approach to case studies and other mentions of your brand.Learn more about Forrester’s predictions here.This post was written by Vice President, Principal Analyst Jeff Pollard and it originally appeared here. More

If you’re using the Rust programming language — or JavaScript, Java, Go or Python — in a project, you may want to check for potential differences between reviewed code versus the compiled code that’s been output. The Rust Security Response working group (WG) has flagged a strange security vulnerability that is being tracked as CVE-2021-42574 and is urging developers to upgrade Rust version 1.56.1. News of the obscure bug was disseminated in a mailing list today. The Rust project has also flagged the Unicode “bidirectional override” issue in a blogpost. But it’s a general bug that doesn’t affect just Rust but all code that’s written in popular languages that use Unicode.  Since it is Unicode, this bug affects not just Rust but other top languages, such as Java, JavaScript, Python, C-based languages and code written in other modern languages, according to security researcher Ross Anderson. Open-source projects such as operating systems often rely on human review of all new code to detect any potentially malicious contributions by volunteers. But the security researchers at Cambridge University said they have discovered ways of manipulating the encoding of source code files so that human viewers and compilers see different logic. “We have discovered ways of manipulating the encoding of source code files so that human viewers and compilers see different logic. One particularly pernicious method uses Unicode directionality override characters to display code as an anagram of its true logic. We’ve verified that this attack works against C, C++, C#, JavaScript, Java, Rust, Go, and Python, and suspect that it will work against most other modern languages,” writes Anderson, detailing this bug and a similar “homoglyph” issue tracked as CVE-2021-42694.”The trick is to use Unicode control characters to reorder tokens in source code at the encoding level. These visually reordered tokens can be used to display logic that, while semantically correct, diverges from the logic presented by the logical ordering of source code tokens. Compilers and interpreters adhere to the logical ordering of source code, not the visual order,” the researchers said. The attack is to use control characters embedded in comments and strings to reorder source code characters in a way that changes its logic.

Software development is international and Unicode — a foundation for text and emoji — supports left-to-right languages, such as English, and right-to-left languages, such as Persian. It does this through “bidirectional override”, an invisible feature called a codepoint that enables embedding left-to-right words inside a right-to-left sentence and vice versa. While they’re normally used to embed a word inside a sentence constructed in the reverse direction, Anderson and Microsoft security researcher Nicholas Boucher discovered that they could be used to change how source code is displayed in certain editors and code review tools. It means that reviewed code can be different than the compiled code and shows how organizations could be hacked through tampered open-source code. “This attack is particularly powerful within the context of software supply chains. If an adversary successfully commits targeted vulnerabilities into open source code by deceiving human reviewers, downstream software will likely inherit the vulnerability,” the researchers warn.Google has found that open-source software supply chain attacks have escalated in the past year. Rust isn’t a widely used programming language, but it has been adopted for systems (versus application) programming by Google, Facebook, Microsoft, Amazon Web Services (AWS) and more for its memory-related safety guarantees. “Rust 1.56.1 introduces two new lints to detect and reject code containing the affected codepoints. Rust 1.0.0 through Rust 1.56.0 do not include such lints, leaving your source code vulnerable to this attack if you do not perform out-of-band checks for the presence of those codepoints,” the Rust project said. The Rust project analyzed its add-on software packages, dubbed “crates” — it reviewed everything published on crates.io from 17 October 2021 — and determined that five crates have the affected codepoints in their source code. However, it didn’t find any malicious codepoints. More

Apple has patched a security flaw in macOS that Microsoft researchers found could be used to install a malicious kernel driver, otherwise known as a ‘rootkit’.  

ZDNet Recommends

The flaw resided within macOS System Integrity Protection (SIP). The glitch allowed a potential attacker to install a hardware interface that allows them to “overwrite system files, or install persistent, undetectable malware”.   The discovery reflects Microsoft’s increased focus on enterprise customers that use a mix of Windows and macOS under hybrid work arrangements, which is evidenced by products like its cross-platform security product, Microsoft Defender for Endpoint. Microsoft introduced Defender ATP for Macs in 2019, well before the pandemic pushed everyone to the hardware they used at home. See also: Ransomware: It’s a ‘golden era’ for cybercriminals – and it could get worse before it gets better. “This OS-level vulnerability and others that will inevitably be uncovered add to the growing number of possible attack vectors for attackers to exploit,” explains Jonathan Bar-Or, from the Microsoft 365 Defender Research team.  “As networks become increasingly heterogeneous, the number of threats that attempt to compromise non-Windows devices also increases.” SIP, aka ‘rootless’, locks down the system from the root by using Apple’s sandbox to protect macOS. It contains several memory-based variables that shouldn’t be able to be modified in non-recovery mode. But SIP can be turned off after booting into recovery mode, allowing an attacker to bypass SIP protections.

“Over the years, Apple has hardened SIP against attacks by improving restrictions,” writes Or.  “One of the most notable SIP restrictions is the filesystem restriction. This is especially important for red teamers and malicious actors, as the amount of damage one can do to a device’s critical components is directly based on their ability to write unrestricted data to disk.” The flaw Microsoft found in Apple’s SIP restrictions was related to system updates, which require unrestricted access to SIP-protected directories. Apple “introduced a particular set of entitlements that bypass SIP checks by design,” writes Or.  Apple patched the flaw, tracked as CVE-2021-30892, in macOS Monterey 12.0.1, as well as updates for Catalina and Big Sur. SIP vulnerabilities aren’t new, but Microsoft decided the bug was serious enough to warrant the name “shrootless”. “While assessing macOS processes entitled to bypass SIP protections, we came across the daemon system_installd, which has the powerful com.apple.rootless.install.inheritable entitlement. With this entitlement, any child process of system_installd would be able to bypass SIP filesystem restrictions altogether,” explains Or.  See also: Cloud security in 2021: A business guide to essential tools and best practices. Microsoft, of course, argues this flaw warrants Defender for Endpoint’s behavioral analytics capabilities to protect Macs in the enterprise.  Apple patched dozens more serious bugs in its latest update for macOS Monterey and earlier. Taking a step back, Microsoft’s post touches on a decades-old debate about whether Macs need antivirus and the two companies’ respective approaches to that question.  Macs, in Apple’s view, don’t need antivirus, whereas Windows PCs do. Apple has used the rise of malware targeting macOS in its arguments against Fortnite-maker Epic Games, for example. And Microsoft this year hired Justin Long, the face of the “Get A Mac” campaigns that once focused on malware targeting Windows PCs but not Macs. But in the enterprise in 2021, where Macs are ascending, work is hybrid, and state-sponsored hackers are looking for every entry point, it’s clear that security threats continue to evolve. More

Image: Getty Images
Signal has released the details of a search warrant it received from police in Santa Clara, California, unveiling the efforts US law enforcement authorities will undertake to force online platforms into disclosing the personal information of their users.In the search warrant, Santa Clara Police sought to get the name, street address, telephone number, and email address of a specific Signal user. It also wanted billing records, the dates of when the account was opened and registered, inbound and outbound call detail records, voicemails, video calls, emails, text messages, IP addresses along with dates and times for each login, and even all dates and times the user connected to Signal.In response to the search warrant, Signal provided law enforcement authorities with timestamps regarding the account specified in the search warrant. The timestamps showed the dates that the account last connected to Signal.Signal said in a blog post that, by default, it does not collect the requested information from users.  “As usual, we couldn’t provide any of that. It’s impossible to turn over data that we never had access to in the first place. Signal doesn’t have access to your messages; your chat list; your groups; your contacts; your stickers; your profile name or avatar; or even the GIFs you search for,” Signal wrote in the blog post.The company’s interaction with Santa Clara County police didn’t end there, however, as the law enforcement authorities then issued a non-disclosure order that required Signal to not publicly disclose that it received the search warrant. The non-disclosure order was then extended four times, which resulted in Signal’s request to unseal the search warrant being repeatedly pushed back. In total, it took Signal almost a full year before the company was able to legally publicly disclose the process it underwent when it received the search warrant.

“Though the judge approved four consecutive non-disclosure orders, the court never acknowledged receipt of our motion to partially unseal, nor scheduled a hearing, and would not return counsel’s phone calls seeking to schedule a hearing,” Signal wrote.Law enforcement authorities around the world are increasingly finding ways to compel online platforms to hand over information about their users. Just last month, hosted email service provider ProtonMail publicly disclosed that French authorities were able to acquire the IP address of one of its users through getting approval from Swiss courts. This was despite ProtonMail not being subject to French or EU requests, and only being required to comply with requests from Swiss authorities.In response to the order, ProtonMail CEO and founder Andy Yen said all companies have to comply with laws, such as court orders, if they operate within 15 miles of land.”No matter what service you use, unless it is based 15 miles offshore in international waters, the company will have to comply with the law,” Yen said at the time.Democracy advocate Freedom House last month also published findings that indicate a growing number of governments are forcing tech businesses to comply with online censorship and surveillance. Freedom House said in the past year alone, 48 out of 70 countries covered in its research — which accounted for 88% of the world’s internet users — have pursued new rules for tech companies on content, data, or competition over the past year.RELATED COVERAGE More