Author: admin

GitHub says npm is among the organizations affected

UAE linked to covert cyber-espionage plot

Voting safeguards based on commuity collateral don’t work if one person can use a momentary loan to “become” 75% of the community.

Exploring what’s next for public-cloud security, including top risks and how to implement better risk management.

Citizen Lab uncovers multi-year campaign targeting autonomous region of Spain, called Catalonia.

In this time of unprecedented cyberwar, organizations must protect the personal digital lives of their executives in order to reduce the company’s risk of direct or collateral damage.

New paper: “Planting Undetectable Backdoors in Machine Learning Models:

Abstract: Given the computational cost and technical expertise required to train machine learning models, users may delegate the task of learning to a service provider. We show how a malicious learner can plant an undetectable backdoor into a classifier. On the surface, such a backdoored classifier behaves normally, but in reality, the learner maintains a mechanism for changing the classification of any input, with only a slight perturbation. Importantly, without the appropriate “backdoor key”, the mechanism is hidden and cannot be detected by any computationally-bounded observer. We demonstrate two frameworks for planting undetectable backdoors, with incomparable guarantees.

First, we show how to plant a backdoor in any model, using digital signature schemes. The construction guarantees that given black-box access to the original model and the backdoored version, it is computationally infeasible to find even a single input where they differ. This property implies that the backdoored model has generalization error comparable with the original model. Second, we demonstrate how to insert undetectable backdoors in models trained using the Random Fourier Features (RFF) learning paradigm or in Random ReLU networks. In this construction, undetectability holds against powerful white-box distinguishers: given a complete description of the network and the training data, no efficient distinguisher can guess whether the model is “clean” or contains a backdoor.

Our construction of undetectable backdoors also sheds light on the related issue of robustness to adversarial examples. In particular, our construction can produce a classifier that is indistinguishable from an “adversarially robust” classifier, but where every input has an adversarial example! In summary, the existence of undetectable backdoors represent a significant theoretical roadblock to certifying adversarial robustness.

Fortinet’s Derek Manky discusses the exponential increase in the speed that attackers weaponize fresh vulnerabilities, where botnets and offensive automation fit in, and the ramifications for security teams.

As an avid internet surfer, you’ve most likely heard of cookies. No, we’re not talking about the ones filled with chocolate chips. We’re talking about the ones that allow you to log in to your favorite websites. Cookies may impact your online security, so check out these tips to manage them and keep your online accounts safe.  

What Are Cookies in Browsers?  

Ever wonder how a website saves the items you placed in your shopping cart last week, even though you closed the tab before making the purchase? This is made possible by cookies. According to the Federal Trade Commission, a cookie is information saved by your web browser. When you visit a website, the site may place a cookie on your web browser so it can recognize your device in the future. If you return to that site later, it can read that cookie to remember you from your last visit, keeping track of your activities over time.1  

First-party vs. Third-party Cookies  

Cookies come in either the first-party or third-party variety. There’s no difference between the two in how they function, but rather in where and how you encountered them. First-party cookies belong to sites you visited first-hand in your browser. Third-party cookies, or “tracking cookies,” generally come from third-party advertising websites. 

Magic Cookies vs. HTTP Cookies 

Although cookies generally function the same, there are technically two different types of cookies. Magic cookies refer to packets of information that are sent and received without changes. Historically, this would be used to log in to a computer database system, such as an internal business  network. This concept predates the modern cookie we use today. 

HTTP cookies are a repurposed version of the magic cookie built for internet browsing and managing online experiences. HTTP cookies help web developers give you more personalized, convenient website experiences. They allow sites to remember you, your website logins, and shopping carts so you can pick back up where you left off from your last visit. However, cybercriminals can manipulate HTTP cookies to spy on your online activity and steal your personal information. 

What Is Cookie Hijacking?  

Cookie hijacking (also known as session hijacking) is typically initiated when a cybercriminal sends you a fake login page. If you click the fake link, the thief can steal the cookie and capture anything you type while on the fraudulent website. Like a phishing attack, cookie hijacking allows a cybercriminal to steal personal information like usernames, passwords, and other important data held within the cookie. If you enter your information while on the fake website, the criminal can then put that cookie in their browser and impersonate you online. They may even change your credentials, locking you out of your account.  

Sometimes, criminals initiate cookie hijacking attacks without a fake link. If you’re browsing on an unsecured, public Wi-Fi connection, hackers can easily steal your data that’s traveling through the connection. This can happen even if the site is secure and your username and password are encrypted. 

Can Cookies Compromise Your Browser Security?  

Because the data in cookies doesn’t change, cookies themselves aren’t harmful. They can’t infect computers with viruses or malware. But if your cookies are hijacked as part of a cyberattack, a criminal could gain access to your browsing history and use cookies as the key to enter your locked accounts. For example, a hacker may steal your identity or confidential company information, purchase items in your online shopping carts, or loot your bank account.  

Tips for a More Secure Browsing Experience 

Preventing cookie hijacking attacks can allow you to browse the internet with greater peace of mind. Follow these tips to not only safeguard your personal information but to also enhance your browsing experience:  

Clean out the cookie jar  

Make it a habit to clear your cookie cache regularly to prevent cookie overload, which could slow your search speeds. Also, almost every browser has the option to enable/disable cookies on your computer. So if you don’t want them at all, your browser’s support section can walk you through how to disable them. 

Turn off autofill features 

Although it’s convenient to not have to re-type your credentials into a website you frequently visit, autofill features could make it easier for a criminal to extract your data with cookie hijacking. Plus, autofill is risky if your physical device falls into the wrong hands. To browse more securely without having to constantly reenter your passwords, use a password manager like McAfee True Key. True Key makes it so you only have to remember one master password, and it encrypts the rest in a vault protected by one of the most secure encryption algorithms available. 

Opt into multi-factor authentication 

Strong, unique passwords for each of your accounts, updated regularly, offer ample protection against hackers. Multi-factor authentication (MFA) adds yet another layer of security by double-checking your identity beyond your username and password, usually with a texted or emailed code. When your accounts offer MFA, always opt in. 

Connect to a virtual private network (VPN) 

Criminals can hijack your cookies if you’re browsing on an unsecured, public Wi-Fi connection. To prevent a criminal from swiping your data, use a virtual private network (VPN), a service that protects your data and privacy online. A VPN creates an encrypted tunnel that makes you anonymous by masking your IP address while connecting to public Wi-Fi hotspots. This is a great way to shield your information from online spies while you’re banking, shopping, or handling any kind of sensitive information online. 

Use antivirus software 

McAfee LiveSafe is an antivirus solution that protects your computer and mobile devices from suspicious web cookies by: 

• Allowing you to keep your online passwords (which are often stored in cookies) in one secure location. 
• Warning you of suspicious links, keeping potentially harmful cookies off your device. 
• Protecting you against viruses and malware. 
• Blocking spam and emails that could lead to sites containing dangerous cookies.  

The post What Are Browser Cookies and How Do I Manage Them? appeared first on McAfee Blog.

Conti — one of the most ruthless and successful Russian ransomware groups — publicly declared during the height of the COVID-19 pandemic that it would refrain from targeting healthcare providers. But new information confirms this pledge was always a lie, and that Conti has launched more than 200 attacks against hospitals and other healthcare facilities since first surfacing in 2018 under its earlier name, “Ryuk.”

On April 13, Microsoft said it executed a legal sneak attack against Zloader, a remote access trojan and malware platform that multiple ransomware groups have used to deploy their malware inside victim networks. More specifically, Microsoft obtained a court order that allowed it to seize 65 domain names that were used to maintain the Zloader botnet.

Microsoft’s civil lawsuit against Zloader names seven “John Does,” essentially seeking information to identify cybercriminals who used Zloader to conduct ransomware attacks. As the company’s complaint notes, some of these John Does were associated with lesser ransomware collectives such as Egregor and Netfilim.

But according to Microsoft and an advisory from the U.S. Cybersecurity & Infrastructure Security Agency (CISA), Zloader had a special relationship with Ryuk/Conti, acting as a preferred distribution platform for deploying Ryuk/Conti ransomware.

Several parties backed Microsoft in its legal efforts against Zloader by filing supporting declarations, including Errol Weiss, a former penetration tester for the U.S. National Security Agency (NSA). Weiss now serves as the chief security officer of the Health Information Sharing & Analysis Center (H-ISAC), an industry group that shares information about cyberattacks against healthcare providers.

Weiss said ransomware attacks from Ryuk/Conti have impacted hundreds of healthcare facilities across the United States, including facilities located in 192 cities and 41 states and the District of Columbia.

“The attacks resulted in the temporary or permanent loss of IT systems that support many of the provider delivery functions in modern hospitals resulting in cancelled surgeries and delayed medical care,” Weiss said in a declaration (PDF) with the U.S. District Court for the Northern District of Georgia.

“Hospitals reported revenue losses due to Ryuk infections of nearly $100 million from data I obtained through interviews with hospital staff, public statements, and media articles,” Weiss wrote. “The Ryuk attacks also caused an estimated $500 million in costs to respond to the attacks – costs that include ransomware payments, digital forensic services, security improvements and upgrading impacted systems plus other expenses.”

The figures cited by Weiss appear highly conservative. A single attack by Ryuk/Conti in May 2021 against Ireland’s Health Service Executive, which operates the country’s public health system, resulted in massive disruptions to healthcare in Ireland. In June 2021, the HSE’s director general said the recovery costs for that attack were likely to exceed USD $600 million.

Conti ravaged the healthcare sector throughout 2020, and leaked internal chats from the Conti ransomware group show the gang had access to more than 400 healthcare facilities in the U.S. alone by October 2020.

On Oct. 28, 2020, KrebsOnSecurity broke the news that FBI and DHS officials had seen reliable intelligence indicating the group planned to ransom many of these care facilities simultaneously. Hours after that October 2020 piece ran, I heard from a respected H-ISAC security professional who questioned whether it was worth getting the public so riled up. The story had been updated multiple times throughout the day, and there were at least five healthcare organizations hit with ransomware within the span of 24 hours.

“I guess it would help if I understood what the baseline is, like how many healthcare organizations get hit with ransomware on average in one week?” I asked the source.

“It’s more like one a day,” the source confided.

A report in February 2022 from Sophos found Conti orchestrated a cyberattack against a Canadian healthcare provider in late 2021. Security software firm Emsisoft found that at least 68 healthcare providers suffered ransomware attacks last year.

While Conti is just one of many ransomware groups threatening the healthcare industry, it seems likely that ransomware attacks on the healthcare sector are underreported. Perhaps this is because a large percentage of victims are paying a ransom demand to keep their data (and news of their breach) confidential. A survey published in February by email security provider Proofpoint found almost 60 percent of victims hit by ransomware paid their extortionists.

Or perhaps it’s because many crime groups have shifted focus away from deploying ransomware and toward stealing data and demanding payment not to publish the information. Conti shames victims who refuse to pay a ransom by posting their internal data on their darkweb blog.

Since the beginning of 2022, Conti has claimed responsibility for hacking a cancer testing lab, a medical prescription service online, a biomedical testing facility, a pharmaceutical company, and a spinal surgery center.

The Healthcare Information and Management Systems Society recently released its 2021 HIMSS Healthcare Cybersecurity Survey (PDF), which interviewed 167 healthcare cybersecurity professionals and found 67 percent had experienced a “significant security incident” in the past year.

The survey also found that just six percent or less of respondent’s information technology budgets were devoted to cybersecurity, although roughly 60 percent of respondents said their cybersecurity budgets would increase in 2022. Last year, just 79 percent of respondents said they’d fully implemented antivirus or other anti-malware systems; only 43 percent reported they’d fully implemented intrusion detection and prevention technologies.

The FBI says Conti typically gains access to victim networks through weaponized malicious email links, attachments, or stolen Remote Desktop Protocol (RDP) credentials, and that it weaponizes Microsoft Office documents with embedded Powershell scripts — initially staging Cobalt Strike via the Office documents and then dropping Emotet onto the network — giving them the ability to deploy ransomware. The FBI said Conti has been observed inside victim networks between four days and three weeks on average before deploying Conti ransomware.