While AI-generation services and major camera makers are adopting the specification for digitally signed metadata, creating a workflow around the nascent ecosystem is still a challenge.

Source: Thapana\_Studio via Shutterstock

The effort to establish digitally verifiable images, video, and other content using signed metadata known as Content Credentials promises an internet where users can easily determine the source of media and whether it is authentic. An easy-to-use implementation of that vision is still under construction.

The brainchild of the Coalition for Content Provenance and Authenticity (C2PA), Content Credentials allow the signing of a metadata audit trail — a manifest — that accompanies images and other media. A photo could be captured with a smartphone, cropped and lightened with photo-editing software, and then compressed by a content delivery network. Content Credentials show each of those steps as part of an audit log, which provides people with information about whether the content they are viewing is real or has been somehow modified. The companies behind the standard include internet providers, software makers, device technology makers, and media giants.

In the last year, major camera makers — Canon, FujiFilm, Leica, Nikon, and Sony — announced support for the Content Credentials standard, but only a few models with the capability have hit the market. Leica has released its second camera with the technology, the Leica SL3-S. At this year's Consumer Electronics Show in January, Samsung announced its upcoming Galaxy S25 mobile phone will support Content Credentials, but only to label images that have been edited by AI. It will not label the original photo taken with the smartphone's camera.

And other major smartphone makers, such as Apple, have not announced support for the standard yet.

Meanwhile, computer-based editing software that supports Content Credentials is mainly limited to Adobe's products. As a founder of the content-authenticity movement, Adobe has extensive support for Content Credentials in its own products — such as a beta feature in Photoshop and tagging images generated with its generative AI service, Adobe Firefly.

Still, some islands of functionality have appeared: Images from OpenAI's DALL-e and Adobe's Firefly are labeled with Content Credentials tagging them as AI-generated, while end-to-end services, such as Truepic, use the technology in its platform to digitally sign and authenticate images.

It's a good start, but an end-to-end workflow requires more: Cameras or smartphones to generate signed images, support for Content Credentials in a wide variety of image-editing software, and the ability to view authenticated metadata on social media and websites.

"It is crucial to ensure that Content Credentials are enabled on all devices throughout the entire workflow—from capture to editing to sharing," says Nico Köhler, head of product experiences for Leica Camera AG. "This requires that all tools and platforms involved in the process support Content Credentials. ... Ensuring compatibility across the entire workflow is key to preserving the authenticity of the content."

**Where Does the Content Credential Work?**

Closed end-to-end ecosystems that incorporate the Content Credentials specification show the potential of the digital authentication feature.

Truepic, for example, essentially created its own end-to-end authenticated image service based on Content Credentials, allowing companies to know the provenance of images. Credit bureaus can request authenticated images of office locations to verify legitimate businesses, dispensing with the need for expensive on-site visits. Insurance companies can request photos of damaged assets from policyholders, avoiding deepfake and edited photos from claiming fictitious damages.

Because smartphones do not yet have the Content Credential technology embedded in the device, Truepic requires its app be used to take photos and video.

"We believe that unless a digital process is backed and secured by image and data authentication, enterprises will not be able to address or even understand the amount of fraud they are missing," says Mounir Ibrahim, chief communications officer at Truepic. "With widespread adoption of credentials, businesses will have a stronger foundation for detecting tampering and ensuring content authenticity, but they will still need a trusted verification platform that can ingest, analyze, and validate digital content at scale and within their environments."

**Interoperable Workflow is Still Not Ready**

While different manufacturers are releasing specific implementations of Content Credentials, creating an end-to-end workflow for images — from creation to editing to distribution to end users — continues to be a challenge.

In early February, for example, Cloudflare tackled one small part of a typical workflow, the content delivery network, announcing that its CDN service would preserve image credentials, even through automated compression and image resizing. Most current infrastructure inadvertently strips away the Content Credential or invalidates the signature by transforming the image.

The Cloudflare changes allow its Images service to add actions to the Content Credential manifest and re-authenticate using the company's credentials, Will Allen, head of privacy and media products at Cloudflare, wrote in a blog post describing the feature.

"When you use Images to resize or change the file format to your images, these transformations will be cryptographically signed by Cloudflare," he said. "This ensures, for example, that the end-user who sees the photograph on your website can use an open-source verification service ... to verify the full provenance chain."

The Content Credential ecosystem at work. A picture taken with a Leica camera is edited with Adobe Photoshop, with each stage documented in the credential. Source: Contentcredential.org/verify using Leica image

Yet, even companies that are forging ahead with their own services look forward to when the wrinkles in the interoperability are ironed out. Truepic's Ibrahim, for example, argues that greater support is the end goal.

"Although closed ecosystems and business processes are where our clients are deriving the most benefit today, we see this as the first step to a more authentic internet," he says. "Interoperable standards will be the essential foundation to a more trusted digital ecosystem and shared global economy."

**About the Author**

Veteran technology journalist of more than 20 years. Former research engineer. Written for more than two dozen publications, including CNET News.com, Dark Reading, MIT's Technology Review, Popular Science, and Wired News. Five awards for journalism, including Best Deadline Journalism (Online) in 2003 for coverage of the Blaster worm. Crunches numbers on various trends using Python and R. Recent reports include analyses of the shortage in cybersecurity workers and annual vulnerability trends.

Content Credentials Show Promise, But Ecosystem Still Young

The open technology, which tackles disinformation, has gained steam in the past year, surpassing 500 corporate members and continuing to evolve.

Source: Tero Vesalainen via Shutterstock

When armed gangs raided a Haitian prison and released 4,700 prisoners last March, images and videos showing the violence and gunfire saturated social media around the world. Determining which graphic media was authentic — and marking those instances in a way that future viewers could verify for themselves what was real and what was modified — was a significant challenge.

Showcasing a solution to exactly that problem, the British Broadcasting Corporation (BBC) launched its adoption of Content Credentials, a technology for attesting to the authenticity and provenance of digital content, by using the technology to verify a TikTok video of the attack. The BBC verified the location of the video and its likely veracity, but it determined that audio of gunfire had been added after the fact. The company then digitally signed the video so that future viewers would know the BBC had verified it.

Content Credentials — an open technology that aims to solve disinformation and media integrity issues — is being developed by the Coalition for Content Provenance and Authenticity (C2PA), a group of more than 500 media, software, and hardware companies working to create an ecosystem for verified media.

The BBC's adoption of Content Credentials is just one use case but an important one, says Andy Parsons, senior director of the Content Authenticity Initiative (CAI) at Adobe and a steering committee member of the C2PA.

"The BBC \[is\] declaring that — regardless of whether this is a photo or whether AI was used to do a photo illustration — you can be guaranteed that it came from BBC, and even that simple proof-of-date or proof-of-origin is something we don't have in media right now," Parsons says.

In 2019, technology and media companies created the CAI to find solutions to disinformation and ways for news organizations to authenticate photos and video. Two years later, six companies — Adobe, Arm, BBC, Intel, Microsoft, and Truepic — founded the C2PA to pursue an open standard for establishing the provenance of various types of media files. The two efforts eventually combined forces to advance Content Credentials, a digital-signature technology and infrastructure for verifying and authenticating media.

In the past year, Content Credentials and the C2PA gained significant steam, with the addition of Amazon, Google, Meta, and OpenAI to the steering committee and the adoption of the technology by several camera makers — including Canon, Leica, and Sony — and smartphone makers, such as Samsung.

Yet the ecosystem is still in its infancy, Christian Paquin, a principal research software engineer at Microsoft, told attendees at last month's ShmooCon 2025.

"You can imagine the situation in five to 10 years when this technology is baked in a lot of the trusted news and a hardware ecosystem that can produce these signatures and can be validated by the social media themselves or the browsers, so that we can really have trust signals to differentiate what's real and what's not," he said.

**Manifest Destiny**

Content Credentials have three main components: the media data, a manifest describing the data and any actions transforming the data, and a digital signature binding the two pieces of information together in a tamper-evident way. The manifest includes typical metadata, as well as some additional C2PA-accepted fields — attestations — that can describe additional attributes of the photo, its source, and the software actions used to process the data.

Based on public-key encryption and digital signatures, Content Credentials act as an audit log of every action performed on a piece of media. A photo could be captured with a smartphone, cropped and lightened with photo-editing software, and then compressed by a content delivery network. Each of those steps would be an attestation in the manifest of the signed content credential.

A video signed with a Content Credential shows that it originated with the BBC News Lab. Source: https://contentcredentials.org/verify

The result is providing an audit trail with each piece of authenticated content, which could help citizens and users better know what is fake and what is arguably real, Adobe's Parsons says.

"All the companies involved — and I would lump in civil society and some governments as well — have a real urgency to solve this problem," he says. "And while C2PA is not a silver bullet at solving misinformation, it does put in place a fundamental foundational layer that the Internet probably should always have had around trust, and this is a really nice clean way to do it."

Already, most makers of foundational artificial intelligence (AI) models — such as Open AI, Meta, and Microsoft — digitally sign all images created by their image-generation models with a Content Credential, indicating that it was AI generated or manipulated.

**An Evolving Standard**

The technology is not done, either. The C2PA specification has quickly evolved over the past three years, reaching version 2.1 in September. Among the new features are efforts to make the credentials more "durable" — in other words, adopting techniques, such as digital fingerprinting and watermarking, to indicate the provenance of images, even if they are manipulated after publication or captured from a screenshot.

The combination of signed metadata with fingerprinting and watermarking is powerful, Adobe's Purdy wrote in a 2024 analysis of the techniques.

"\[N\]one of these techniques is durable enough in isolation to be effective on its own," he said. "But combined into a single approach, the three form a unified solution that is robust and secure enough to ensure that reliable provenance information is available no matter where a piece of content goes."

Watermarks, fingerprinting, and Content Credentials can be a powerful combination. Source: "Durable Content Credentials," CAI blog

A number of technical problems remain to be solved. Journalists reporting on an authoritarian regime, for example, may want to remain anonymous and mask their location. Zero-knowledge proofs can help, allowing a name to be redacted and only the organization — BBC News, for example — to be shown or to generalize the location. ZK proofs allow an attribute to be signed, so a photo's GPS coordinates could instead be reduced to New York City or Kiev to provide a general location that hides the photographer's identity, and verified with a digital signature.

"This is an evolving set of certifications," Microsoft's Paquin told the audience at ShmooCon. "The main challenge is establishing who can sign certificates ... establishing PKI that is worldwide is a big problem."

**About the Author**

Veteran technology journalist of more than 20 years. Former research engineer. Written for more than two dozen publications, including CNET News.com, Dark Reading, MIT's Technology Review, Popular Science, and Wired News. Five awards for journalism, including Best Deadline Journalism (Online) in 2003 for coverage of the Blaster worm. Crunches numbers on various trends using Python and R. Recent reports include analyses of the shortage in cybersecurity workers and annual vulnerability trends.

Content Credentials Technology Verifies Image, Video Authenticity

PRESS RELEASE

LONDON, Jan. 20, 2025 /PRNewswire/ -- A new survey from Omdia reveals that phishing scams are the leading security threat for smartphone users, with 24% of respondents reporting they have fallen victim to these attacks. Phishing, which includes fraudulent texts, emails, or calls designed to trick individuals into revealing sensitive personal information, remains a significant concern as cybercriminals continue to exploit unsuspecting consumers.

Omdia surveyed 1,572 consumers across the Americas, Asia & Oceania, and Europe in October 2024 for the fourth annual Omdia Mobile Device Security Scorecard. The survey found that the next most common security issue was malware and viruses, followed by physical theft, such as pickpocketing, mugging, or snatching.

In Omdia's recent assessment of leading premium smartphones, Google's Pixel 9 Pro and Samsung's Galaxy S24 outperformed Apple's iPhone 16 Pro and other Android-based devices, including the OnePlus 12, Xiaomi 14, and Honor Magic 6 Pro. Anti-phishing protection proved to be a weak spot across all devices, as none successfully intercepted all phishing texts, calls and emails.  

Simulated spam calls revealed that all Android devices from Google, Xiaomi, OnePlus, Honor, and Samsung successfully flagged suspected spam calls before users answered, while the iPhone 16 Pro lacked similar voice call protection. 

None of the tested devices fully No device flagged simulated phishing emails from Gmail as phishing, only identifying them as spam when sent from Google's SMTP. 

Despite gaps in detecting phishing texts and emails, devices with Google Safe Browsing protections successfully blocked the link from opening, displaying a warning screen and requiring user confirmation to proceed. Performance across browsers varied significantly: Samsung Internet effectively blocked most links, including advanced custom URLs, while Xiaomi Mii and OnePlus Internet browsers failed to warn users about known malicious links, underscoring inconsistencies in Android device's security.

"The lack of security protection, particularly against the growing threat of phishing attacks, is eroding consumer trust," said Omdia Senior Analyst, Aaron West. "When consumers were asked if their trust following a security issue increased (due to how well the issue was handled) or decreased, 73% reported they had reduced trust in the smartphone brand and operating system developer."  

"Despite the latest protections in place by some manufacturers, it is difficult to protect 100% against phishing attempts, highlighting the severity of the issue and potential impact to consumers. That said, smartphone manufacturers can (demonstrated by the more advanced phishing protection capabilities available) and should have a better baseline of phishing protection – such as voice call protection, and all Android devices making use of Google's Safe Browsing protections," said Hollie Hennessy, Principal Analyst, Omdia. "This needs to be paired with awareness activity from manufacturers and the wider industry to help consumers be vigilant and prepared".

ABOUT OMDIA

Omdia, part of Informa TechTarget, Inc. (Nasdaq: TTGT), is a technology research and advisory group. Our deep knowledge of tech markets combined with our actionable insights empower organizations to make smart growth decisions.

Omdia Finds Phishing Attacks Top Smartphone Security Concern for Consumers

While employees want to take advantage of the increased efficiency of GenAI and LLMs, CISOs and IT teams must be diligent and stay on top of the most up-to-date security regulations.

Anuj Jaiswal, Chief Product Officer, Fortanix

January 23, 2025

3 Min Read

Source: LuckyStep48 via Alamy Stock Vector

COMMENTARY

The rapid rise of artificial intelligence (AI) has cast a long shadow, but its immense promise comes with a significant risk: shadow AI.

Shadow AI refers to the use of AI technologies, including AI models and generative AI (GenAI) tools outside of a company's IT-sanctioned governance. As more people use tools like ChatGPT to increase their efficiency at work, many organizations are banning publicly available GenAI for internal use. Among the organizations looking to prevent unnecessary security risks are those in the financial services and healthcare sectors, as well as technology companies like Apple, Amazon, and Samsung.

Unfortunately, enforcing such a policy is an uphill battle. According to a recent report, non-corporate accounts make up 74% of ChatGPT use and 74% of Gemini and Bard use at work. Employees can easily skirt corporate policies to continue their AI use for work, potentially opening up security risks.

The greatest among these is the lack of protection for sensitive data. As of March 2024, 27.4% of data inputted into AI tools would be considered sensitive, an increase from 10.7% at the same time last year. Protecting this information once it is put into a GenAI tool is virtually impossible.

The uncontrolled risk of shadow AI usage reveals the need for stringent privacy and security practices when employees use AI.

It all boils down to data. Data is the fuel of AI, but it is also the most valuable asset to organizations. Stolen, leaked, or corrupted data causes real, tangible harm to a business — regulatory fines from leaking personally identifiable information (PII), costs associated with leaked proprietary information like source code, and an increase in severe security breaches like hacks and malware.

To mitigate risk, organizations must secure their data while it's at rest, in transit, and in use. The counter to risky shadow AI use is having fine control over the information employees feed into large language models (LLMs).

**How Can CISOs Secure GenAI and Company Data?**

Securing sensitive company data is a challenging balancing act for chief information security officers (CISOs) as they weigh the desire for their organizations to take advantage of the perceived value of GenAI while also protecting the sole asset that makes these benefits possible — their data.

So, the question becomes: How do you do this? How do you get the balance right? How do you extract positive business outcomes while protecting the enterprise's most valuable asset?

At a high level, CISOs should look at protecting data through its entire life cycle. This includes:

*   Protecting the data before it is even ingested into the GenAI model
    

*   Ensuring that the data output is completely secured, as this new data will drive the business outcomes and create true value
    

If the data life cycle isn't secure, this becomes a business-critical exposure.

More specifically, a multifaceted approach is necessary to protect sensitive data from being leaked, and though it starts with limiting shadow AI as much as possible, it is just as important to preserve data security and privacy with some basic best practices:

*   Encryption: Data encryption across its life cycle is vital, but it's equally important to manage and store encryption keys securely and separately from the data itself.
    
*   Obfuscation: Use data tokenization to anonymize any sensitive or PII data that could be fed to an LLM. This prevents data that enters the AI pipeline from being corrupted or leaked.
    
*   Access: Apply granular, role-based access controls to data so that only authorized users can see and use the data in plain text.
    
*   Governance: Commit to ethical business practices, embed data privacy across all operations, and remain current on data privacy regulations.
    

As is often the case with most tech advancements, GenAI's ease and convenience come with some fallbacks. While employees want to take advantage of the increased efficiency of GenAI and LLMs for work, CISOs and IT teams must be diligent and stay on top of the most up-to-date security regulations to prevent sensitive data from entering the AI system. Along with making sure workers know the importance of data protection, it is key to mitigate potential risks by taking all measures to encrypt and secure data from the start.

**About the Author**

Chief Product Officer, Fortanix

Anuj Jaiswal is chief product officer at Fortanixa. He is seasoned product and engineering leader with an impressive track record spanning more than 20 years. Anuj previously served as vice president of engineering at Embrace.io and was a pivotal figure at Arkin, a company that caught the attention of VMware and was subsequently acquired in 2016. During his time at Arkin, Anuj played a critical role in the acquisition process, showcasing his strategic vision and technical insight. Before Arkin, Anuj worked as a senior leader at FrontRange Solutions (acquired by Ivanti) and DBO2 (acquired by Predictive Solutions).

Anuj's domain knowledge is extensive, covering crucial areas such as cloud security, network security, cybersecurity, data security and application performance monitoring (APM) for both cloud and mobile applications. His approach to product management — vision, strategy, customer-centric, roadmapping, partnerships, and UX — has consistently resulted in products that drive revenue and growth for organizations while meeting customer needs. Anuj holds a master's degree in computer applications and a bachelor's degree in business administration.

The Security Risk of Rampant Shadow AI

New hands-on testing results show that most devices are unable to catch phishing emails, texts, or calls, leaving users at risk.

Source: Thomas Bethge via Alamy Stock Photo

Highlights:

*   Consumer survey shows that the most common security issues are phishing attacks, such as text or emails seeking personal information, as well as malware and physical theft. 
    

*   Hands-on device testing shows that Samsung S24 offers the best anti-phishing protection, while the Google Pixel 9 Pro leads in many other areas. 
    

*   The iPhone 16 Pro and other premium Android smartphones from Honor, Xiaomi, and OnePlus lack security features and protections. 
    

COMMENTARY  
New research from Omdia shows that the security capabilities on several of the latest consumer smartphones, including top devices from Apple, Google, and Samsung, fail to detect several types of common phishing attacks.

As part of the fourth-annual Omdia Mobile Device Security Scorecard, Omdia surveyed 1,572 consumers across 13 major countries in the Americas, Asia and Oceania, and Europe, in October 2024. This survey covered the demographics of smartphone users, their security concerns and attitudes, their perception of the most common security threats, and the key smartphone purchasing drivers.

Consumers reported that the most common security issue they encountered was phishing scams and attacks (texts, emails or calls which use false pretenses to try to get the target to give away valuable personal information), with 24% reporting to have experienced phishing. 

Omdia also asked consumers to rate how important various security features were, with anti-phishing having the largest net-importance rating. In fact, over the four years Omdia has conducted its Mobile Device Security Scorecard research, anti-phishing has been rising in importance. This is no surprise, as phishing not only becomes more common, but also as consumers become more familiar with the attacks and the potential for personal information theft or financial loss. 

The next most common security issue was malware and viruses. The third most common was physical theft, such as pickpocketing, mugging or snatching. 

**Mobile Device Security Testing Results**

For each of these security issues, Omdia tested leading premium smartphones to determine availability and effectiveness of security features and capabilities. Google's Pixel 9 Pro and Samsung's Galaxy S24 both scored highly, ahead of Apple's iPhone 16 Pro and other leading Android-based devices, including the OnePlus 12, Xiaomi 14, and Honor Magic 6 Pro.

Despite the importance consumers place on phishing prevention on mobile devices, the anti-phishing features failed to some degree on every device Omdia tested. No device caught all the attempted phishing texts, calls, and emails initiated as part of the testing.  

Several of the devices did catch simulated spam calls: All Android devices from Google, Xiaomi, OnePlus, Honor, and Samsung, which have voice call protection, flagged suspected spam calls before the recipient could answer the call. The iPhone 16 Pro, however, did not have the same protections and did not catch the simulated spam call. 

Text messages with malicious short links were sent from an unknown number and sender ID, but these were not successfully caught by the test devices. Simulated phishing emails were sent from both Gmail and Google's Mail Delivery subsystem, but no device caught the phishing emails from Gmail; the messages were only identified as spam when sent from Google's SMTP. 

Despite not all phishing texts and emails being caught, once malicious links were opened on the devices, those that use Google Safe Browsing protections successfully blocked the link from opening. A warning screen was raised, forcing the user to bypass it to proceed.

Yet not all Android devices performed similarly: Samsung Internet blocked most links except the more sophisticated custom URLs, while the Xiaomi Mii and OnePlus Internet browsers failed to warn the user even when loading known malicious links. 

In previous years' testing, smartphones have been able to detect and successfully block or warn against the attempted phish. However, the difference year-on-year showcases the ever-changing nature of threats.

**Lack of Mobile Security Impacts Consumer Trust**

The lack of effective security protections on mobile devices, particularly against the growing threat of phishing attacks, is eroding consumer trust.

When Omdia asked consumers if their trust following a security issue increased (due to how well the issue was handled) or decreased, 73% reported they had reduced trust in the smartphone brand and operating system developer. 

Despite some manufacturers' efforts to ensure the latest mobile device security protections are in place, Omdia notes that it is difficult to protect against 100% of phishing attempts. This highlights the severity of the issue and potential impact to consumers.

That said, Omdia asserts that smartphone manufacturers can (demonstrated by the more advanced phishing protection capabilities on the market) and should have more effective baseline phishing protection in place — such as voice call protection, and all Android devices making use of Google’s Safe Browsing protections. 

Omdia sees the value in the phishing protection features implemented by the leading smartphone vendors, and how this can help to protect consumers — at least in most instances, based on Omdia's latest round of testing.

However, these features need to be paired with awareness activity from manufacturers, as well as the wider industry, to help consumers be vigilant and prepared for the instance when a security threat evades a protection mechanism. Such awareness adds an important additional layer of protection against the rising number of scams targeting consumers through their mobile devices.

**About the Authors**

Principal Analyst, IoT Cybersecurity, Omdia

Hollie works as a Principal Analyst within the Omdia team, providing insight into the fascinating and fast-moving domain of IoT Cybersecurity.

Hollie has a range of experience in research. She began her career in the legal sector, writing and researching for expert witness reports on the labour market. She then moved into product testing, with a consumer protection focus. In this role she was responsible for managing comparative tests of various tech products, as well as regular testing and investigative work into the security of these products.

She has published various articles for Which?, the UK's largest consumer organisation and one of the largest subscription magazines, and Computing magazine.

Senior Analyst, Smartphones, Omdia

Aaron West provides expertise and analysis of the smartphone market, joining Omdia in 2022. His areas of reporting include the active smartphone installed base, eSIM development, mmWave 5G, sustainability, and the foldable smartphone market.

Aaron has a range of research and publishing experience. He graduated with a degree in theoretical physics before moving into consumer product testing, with a focus on sustainability and the environment.

In this role, he was responsible for research project management, including comparative testing of technology products and consumer surveys. He also reported on the energy efficiency and longevity of home appliances, as well as the product lifecycle and obsolescence of smartphones.

Phishing Attacks Are the Most Common Smartphone Security Issue for Consumers

Cybersecurity researchers have detailed a now-patched security flaw impacting Monkey's Audio (APE) decoder on Samsung smartphones that could lead to code execution.
The high-severity vulnerability, tracked as CVE-2024-49415 (CVSS score: 8.1), affects Samsung devices running Android versions 12, 13, and 14.
"Out-of-bounds write in libsaped.so prior to SMR Dec-2024 Release 1 allows remote

Google Project Zero Researcher Uncovers Zero-Click Exploit Targeting Samsung Devices

A free VPN app called Big Mama is selling access to people’s home internet networks. Kids are using it to cheat in a VR game while researchers warn of bigger security risks.

In the hit virtual reality game _Gorilla Tag_, you swing your arms to pull your primate character around—clambering through virtual worlds, climbing up trees and, above all, trying to avoid an infectious mob of other gamers. If you’re caught, you join the horde. However, some kids playing the game claim to have found a way to cheat and easily “tag” opponents.

Over the past year, teenagers have produced video tutorials showing how to side-load a virtual private network (VPN) onto Meta’s virtual reality headsets and use the location-changing technology to get ahead in the game. Using a VPN, according to the tutorials, introduces a delay that makes it easier to sneak up and tag other players.

While the workaround is likely to be an annoying but relatively harmless bit of in-game cheating, there’s a catch. The free VPN app that the video tutorials point to, Big Mama VPN, is also selling access to its users’ home internet connections—with buyers essentially piggybacking on the VR headset’s IP address to hide their own online activity.

This technique of rerouting traffic, which is best known as a residential proxy and more commonly happens through phones, has become increasingly popular with cybercriminals who use proxy networks to conduct cyberattacks and use botnets. While the Big Mama VPN works as it is supposed to, the company’s associated proxy services have been heavily touted on cybercrime forums and publicly linked to at least one cyberattack.

Researchers at cybersecurity company Trend Micro first spotted Meta’s VR headsets appearing in its threat intelligence residential proxy data earlier this year, before tracking down that teenagers were using Big Mama to play _Gorilla Tag_. An unpublished analysis that Trend Micro shared with WIRED says its data shows that the VR headsets were the third most popular devices using the Big Mama VPN app, after devices from Samsung and Xiaomi.

“If you’ve downloaded it, there’s a very high likelihood that your device is for sale in the marketplace for Big Mama,” says Stephen Hilt, a senior threat researcher at Trend Micro. Hilt says that while Big Mama VPN may be being used because it is free, doesn’t require users to create an account, and apparently doesn’t have any data limits, security researchers have long warned that using free VPNs can open people up to privacy and security risks.

These risks may be amplified when that app is linked to a residential proxy. Proxies can “allow people with malicious intent to use your internet connection to potentially use it for their attacks, meaning that your device and your home IP address may be involved in a cyberattack against a corporation or a nation state,” Hilt says.

"Gorilla Tag is a place to have fun with your friends and be playful and creative—anything that disturbs that is not cool with us,” a spokesperson for Gorilla Tag creator Another Axiom says, adding they use “anti-cheat mechanisms” to detect suspicious behavior. Meta did not respond to a request for comment about VPNs being side-loaded onto its headsets.

**Proxies Rising**

Big Mama is made up of two parts: There’s the free VPN app, which is available on the Google Play store for Android devices and has been downloaded more than 1 million times. Then there’s the Big Mama Proxy Network, which allows people (among other options) to buy shared access to “real” 4G and home Wi-Fi IP addresses for as little as 40 cents for 24 hours.

Vincent Hinderer, a cyber threat intelligence team manager who has researched the wider residential proxy market at Orange Cyberdefense, says there are various scenarios where residential proxies are used, both for people who are having traffic routed through their devices and also those buying and selling proxy services. “It’s sometimes a gray zone legally and ethically,” Hinderer says.

For proxy networks, Hinderer says, one end of the spectrum is where networks could be used as a way for companies to scrape pricing details from their competitors' websites. Other uses can include ad verification or people scalping sneakers during sales. They may be considered ethically murky but not necessarily illegal.

At the other end of the scale, according to Orange’s research, residential proxy networks have broadly been used for cyber espionage by Russian hackers, in social engineering efforts, as part of DDoS attacks, phishing, botnets, and more. “We have cybercriminals using them knowingly,” Hinderer says of residential proxy networks generally, with Orange Cyberdefense having frequently seen proxy traffic in logs linked to cyberattacks it has investigated. Orange’s research did not specifically look at uses of Big Mama's services.

Some people can consent to having their devices used in proxy networks and be paid for their connections, Hinderer says, while others may be included because they agreed to it in a service’s terms and conditions—something research has long shown people don’t often read or understand.

Big Mama doesn’t make it a secret that people who use its VPN will have other traffic routed through their networks. Within the app it says it “may transport other customer’s traffic through” the device that’s connected to the VPN, while it is also mentioned in the terms of use and on a FAQ page about how the app is free.

The Big Mama Network page advertises its proxies as being available to be used for ad verification, buying online tickets, price comparison, web scraping, SEO, and a host of other use cases. When a user signs up, they’re shown a list of locations proxy devices are located in, their internet service provider, and how much each connection costs.

This marketplace, at the time of writing, lists 21,000 IP addresses for sale in the United Arab Emirates, 4,000 in the US, and tens to hundreds of other IP addresses in a host of other countries. Payments can only be made in cryptocurrency. Its terms of service say the network is only provided for “legal purposes,” and people using it for fraud or other illicit activities will be banned.

Despite this, cybercriminals appear to have taken a keen interest in the service. Trend Micro’s analysis claims Big Mama has been regularly promoted on underground forums where cybercriminals discuss buying tools for malicious purposes. The posts started in 2020. Similarly, Israeli security firm Kela has found more than 1,000 posts relating to the Big Mama proxy network across 40 different forums and Telegram channels.

Kela’s analysis, shared with WIRED, shows accounts called “bigmama\_network” and “bigmama” posted across at least 10 forums, including cybercrime forums such as WWHClub, Exploit, and Carder. The ads list prices, free trials, and the Telegram and other contact details of Big Mama.

It is unclear who made these posts, and Big Mama tells WIRED that it does not advertise.

Posts from these accounts also said, among other things, that “anonymous” bitcoin payments are available. The majority of the posts, Kela’s analysis says, were made by the accounts around 2020 and 2021. Although, an account called “bigmama\_network” has been posting on the clearweb Blackhat World SEO forum until October this year, where it has claimed its Telegram account has been deleted multiple times.

In other posts during the last year, according to the Kela analysis, cybercrime forum users have recommended Big Mama or shared tips about the configurations people should use. In April this year, security company Cisco Talos said it had seen traffic from the Big Mama Proxy, alongside other proxies, being used by attackers trying to brute force their way into a variety of company systems.

**Mixed Messages**

Big Mama has few details about its ownership or leadership on its website. The company’s terms of service say that a business called BigMama SRL is registered in Romania, although a previous version of its website from 2022, and at least one live page now, lists a legal address for BigMama LLC in Wyoming. The US-based business was dissolved in April and is now listed as inactive, according to the Wyoming Secretary of State’s website.

A person using the name Alex A responded to an email from WIRED about how Big Mama operates. In the email, they say that information about free users’ connections being sold to third parties through the Big Mama Network is “duplicated on the app market and in the application itself several times,” and people have to accept the terms of conditions to use the VPN. They say the Big Mama VPN is officially only available from the Google Play Store.

“We do not advertise and have never advertised our services on the forums you have mentioned,” the email says. They say they were not aware of the April findings from Talos about its network being used as part of a cyberattack. “We do block spam, DDOS, SSH as well as local network etc. We log user activity to cooperate with law enforcement agencies,” the email says.

The Alex A persona asked WIRED to send it more details about the adverts on cybercrime forums, details about the Talos findings, and information about teenagers using Big Mama on Oculus devices, saying they would be “happy” to answer further questions. However, they did not respond to any further emails with additional details about the research findings and questions about their security measures, whether they believe someone was impersonating Big Mama to post on cybercrime forums, the identity of Alex A, or who runs the company.

During its analysis, Trend Micro’s Hilt says that the company also found a security vulnerability within the Big Mama VPN, which could have allowed a proxy user to access someone’s local network if exploited. The company says it reported the flaw to Big Mama, which fixed it within a week, a detail Alex A confirmed.

Ultimately, Hilt says, there are potential risks whenever anyone downloads and uses a free VPN. “All free VPNs come with a trade-off of privacy or security concerns,” he says. That applies to people side-loading them onto their VR headsets. “If you’re downloading applications from the internet that aren't from the official stores, there’s always the inherent risk that it isn’t what you think it is. And that comes true even with Oculus devices.”

This VPN Lets Anyone Use Your Internet Connection. What Could Go Wrong?

Large language models (LLMs) can help app security firms find and fix software vulnerabilities. Malicious actors are on to them too, but here's why defenders may retain the edge.

Source: vs148 via Shutterstock

Security researchers and attackers are turning to AI models to find vulnerabilities, a technology whose use will likely drive the annual count of software flaws higher, but could eventually result in fewer flaws in public releases, experts say.

On Nov. 1, Google said its Big Sleep large language model (LLM) agent discovered a buffer-underflow vulnerability in the popular database engine, SQLite. The experiment shows both the peril and the promise of AI-powered vulnerability discovery tools: The AI agent searched through the code for variations on a specific vulnerability, but identified the software flaw in time for Google to notify the SQLite project and work with them to fix the issue.

Using AI just for software-defect discovery could result in a surge in vulnerability disclosures, but introducing LLM agents into the development pipeline could reverse the trend and lead to fewer software flaws escaping into the wild, says Tim Willis, head of Google's Project Zero, the company's effort to identify zero-day vulnerabilities.

"While we are at an early stage, we believe that the techniques we develop through this research will become a useful and general part of the toolbox that software developers have at their disposal," he says.

Google is not alone in searching for better ways to find — and fix — vulnerabilities. In August, a group of researchers from Georgia Tech, Samsung Research, and other firms — collectively known as Team Atlanta — used an LLM bug-finding system to automatically find and patch a bug in SQLite. And just last month, cybersecurity firm GreyNoise Intelligence revealed it had used its Sift AI system to analyze honeypot logs leading to the discovery and patching of two zero-day vulnerabilities affecting Internet-connected cameras used in sensitive environments.

Overall, companies are gaining more ways to automate vulnerability discovery, and — if they are serious about security — will be able to drive down the number of vulnerabilities in their products by using the tools in development, says Corey Bodzin, chief product officer at GreyNoise Intelligence.

"The exciting thing is we do have technology that allows people who \[care about\] security to be more effective," he says. "Sadly ... there are not many companies where that is ... a primary driver, but even in companies where \[security is\] purely viewed as a cost" can benefit from using these tools.

**Only the First Steps**

Currently, Google's custom approach is still bespoke and requires work to adapt to specific vulnerability-finding tasks. The company's Big Sleep agent does not to look for completely new vulnerabilities, but uses details from a previously discovered vulnerability to look for similar issues. The project has looked at smaller programs with known vulnerabilities as test cases, but the SQLite experiment is the first time they found vulnerabilities in production code, the Google Project Zero and Google DeepMind researchers stated in Google's blog post describing the research.

While specialized fuzzers would likely have found the bug, tuning those tools to perform well is a very manual process, says Google's Willis.

"One promise of \[L\]LM agents is that they might generalize across applications without the need for specialized tuning," he says. "Additionally, we're hopeful that \[L\]LM agents will be able to uncover a different subset of vulnerabilities than those typically found through fuzzing."

The use of AI-based vulnerability discovery tools will be a race between attackers and defenders. Manual code review is a viable way of finding bugs for attackers, who only need a single exploitable vulnerability or short chain of vulnerabilities. But defenders need a scalable way of finding and fixing applications, Willis says. While bug-finding tools can be a force multiplier for both attackers and defenders, the ability to scale up to analyze code will likely be a greater benefit for defenders, Willis says.

"We expect that advances in automated vulnerability discovery, triage, and remediation will disproportionately benefit defenders," he says.

**Focus AI on Finding and Fixing Bugs**

Companies that focus on using AI to generate secure code and fix bugs when found will deliver higher quality code from developers, says Chris Wysopal, co-founder and chief security evangelist at Veracode, an application security firm. He argues that automating bug finding and bug fixing are two completely different problems. Finding vulnerabilities is a very large data problem, whIle fixing bugs usually deals with perhaps a dozen lines of code.

"Once you know the bug is there — if you found it through fuzzing, or through an LLM, or using human code review — and you know what kind of bug it is, fixing it is relatively easy," he says. "So, LLMs favor defenders, because having access to source code and fixing issues is easy. So I'm kind of bullish that we can eliminate whole classes of vulnerabilities, but it's not from finding more, it's from being able to fix more."

Companies that require developers to run automated security tools before code check-in will find themselves on a path to paying down their security debt — the collection of issues that they know about, but have not had time to fix, he says. Currently, about half (46%) of organizations have security debt in the form of persistent critical flaws in applications, according to Veracode's 2024 State of Software Security report.

"The idea that you're committing code that has a problem in it, and it's not fixed, will become the exception, not the rule, like it is today," Wysopal says. "Once you can start to automate this fixing — and we're always getting better at automating finding \[vulnerabilities\] — I think that's how things change."

Yet, the technology will still have to overcome companies' focus on efficiency and productivity over security, says Bob Rudis, vice president of data science and security research at GreyNoise Intelligence. He points to the fixing of the two security vulnerabilities that GreyNoise Intelligence found and responsibly disclosed. The company only fixed the issues in two product models, but not others — despite the fact that the other products likely had similar issues, he says.

Google and GreyNoise Intelligence proved that the technology will work, but whether companies integrate AI into the development pipelines to eliminate bugs is still an open question.

Rudis has doubts.

"I'm sure a handful of organizations are going to deploy it — it's going to make like seven C files a little bit safer across a bunch of organizations, and maybe we'll get like a tick more security for the ones that can actually deploy it properly," he says. "But ultimately, until we actually change the incentive structure around how software vendors build and deploy things, and how consumers actually purchase and deploy and configure things, we are not going to see any benefit."

**About the Author**

Veteran technology journalist of more than 20 years. Former research engineer. Written for more than two dozen publications, including CNET News.com, Dark Reading, MIT's Technology Review, Popular Science, and Wired News. Five awards for journalism, including Best Deadline Journalism (Online) in 2003 for coverage of the Blaster worm. Crunches numbers on various trends using Python and R. Recent reports include analyses of the shortage in cybersecurity workers and annual vulnerability trends.

AI &amp; LLMs Show Promise in Squashing Software Bugs

IBM Security Verify Access versions prior to 10.0.8 suffer from authentication bypass, reuse of private keys, local privilege escalation, weak settings, outdated libraries, missing password, hardcoded secrets, remote code execution, missing authentication, null pointer dereference, and lack of privilege separation vulnerabilities.

IBM Security Verify Access 32 Vulnerabilities

The prominent state-sponsored advanced persistent threat (APT), aka Jumpy Pisces, appears to be moving away from its primary cyber-espionage motives and toward wreaking widespread disruption and damage.

Source: DD Images via Shutterstock

One of North Korea's most prominent state-sponsored threat groups has pivoted to using Play ransomware in recent attacks, signifying the first time the group has partnered up with an underground ransomware network. Worryingly, it sets the stage for future high-impact attacks, researchers surmise.

According to Palo Alto Networks' Unit 42, which tracks the advanced persistent threat (APT) as Jumpy Pisces (aka Nickel Hyatt, Onyx Sleet, Silent Chollima, Stonefly, and TDrop2), Andariel is now working with the Play ransomware gang, but whether it's as an initial access broker (IAB) or affiliate of the ransomware group is not clear, the researchers observed in a blog post on Oct. 31. Previously, Andariel was associated with a ransomware strain called "Maui" that's been active since at least 2022.

Unit 42 researchers believe the group is responsible for a Play ransomware attack discovered last month in which attackers gained initial access to a network via a compromised user account several months before, in May. Andariel moved laterally after its initial network breach and maintained persistence by spreading the open source tool Sliver and its unique custom malware, DTrack, to other hosts via the Server Message Block (SMB) protocol, according to Unit 42. Months later, in early September, it deployed the Play payload.

Related:Cybersecurity Training Resources Often Limited to Developers

"This shift in their tactics, techniques and procedures (TTPs) signals deeper involvement in the broader ransomware threat landscape," Unit 42 researchers wrote in the post. "This development could indicate a future trend where North Korean threat groups will increasingly participate in broader ransomware campaigns, potentially leading to more widespread and damaging attacks globally."

**Ransomware in Transition?**

Play ransomware, maintained and deployed by a group tracked as Fiddling Scorpius, made its claim to fame by targeting the city of Oakland, Calif., in February 2023 with a crippling attack. It then quickly rose up the threat ranks to become a major player in the game.

Some researchers have suggested that Fiddling Scorpius has transitioned from mounting its own attacks to a ransomware-as-a-service (RaaS) model, according to Unit 42. However, the group itself has announced on its Play ransomware leak site that it does not provide a RaaS ecosystem, according to the researchers. If this is true, then Andariel most likely acted as an IAB in the attack rather than an affiliate, they said.

Either way, "network defenders should view … \[the\] activity as a potential precursor to ransomware attacks, not just espionage, underscoring the need for heightened vigilance," according to Unit 42.

Related:Codasip Donates Tools to Develop Memory-Safe Chips

There were several clues in the attack sequence that point to collaboration between Andariel and the Play ransomware. For one, the compromised account that attackers used for initial access and subsequent spreading of Andariel's signature tools, including Silver and Dtrack, was the same one used prior to ransomware deployment.

"The ransomware actor leveraged the account to abuse Windows access tokens, move laterally and escalate to SYSTEM privileges via PsExec," according to the post. "This eventually led to the mass uninstallation of endpoint detection and response (EDR) sensors and the onset of Play ransomware activity."

The researchers also observed command-and-control (C2) communication with the Silver malware the day before Play ransomware was deployed. Moreover, Play ransomware attacks are known for leaving tools in the in the folder C:\\Users\\Public\\Music, and some tools used prior to ransomware deployment in the Andariel attack also were located there, the researchers noted.

**Defenders Beware Rising North Korean Ransomware Threat**

Andariel has been active for several years and has mounted a number of high-profile attacks that have targeted critical defense, aerospace, nuclear, and engineering companies as well as global managed service providers.

Related:Samsung Zero-Day Vuln Under Active Exploit, Google Warns

Andariel is controlled by North Korea's military intelligence agency, the Reconnaissance General Bureau, which is involved in the nation's illicit arms trade and responsible for its malicious cyber activity. The group's antics already have drawn the attention of international law enforcement, including the US National Security Agency (NSA), which considers the group an ongoing threat to various industry sectors, particularly in the US, South Korea, Japan, and India.

The US Department of State's Rewards for Justice (RFJ) is even offering a reward of up to $10 million for information that could lead it to Rim Jong Hyok, a key player in Andariel's management structure, or any co-conspirators in the group.

Given the need for worldwide organizations to be on alert, Unit 42 included a list of indicators of compromise (IoCs) in its blog post. The researchers advised that defenders leverage the latest threat intelligence to identify malware on networks, and advanced URL filtering and DNS security products to spot known URLs and domains associated with Andariel's malicious activity.

**About the Author**

Elizabeth Montalbano is a freelance writer, journalist, and therapeutic writing mentor with more than 25 years of professional experience. Her areas of expertise include technology, business, and culture. Elizabeth previously lived and worked as a full-time journalist in Phoenix, San Francisco, and New York City; she currently resides in a village on the southwest coast of Portugal. In her free time, she enjoys surfing, hiking with her dogs, traveling, playing music, yoga, and cooking.

Tag

#samsung