Emerging Threats & Vulnerabilities to Prepare for in 2025

In 2024, we at Dark Reading covered a variety of attacks, exploits, and, of course, vulnerabilities across the board. Here, we recount 10 emerging threats organizations should be prepared for — as detailed by Dr. Jason Clark in "10 Emerging Vulnerabilities Every Enterprise Should Know," a Dark Reading webinar — as they continuously rise and develop in 2025.

Zero-Day Exploits

Zero-days and their increase in volume across the cybersecurity landscape is a particularly concerning trend, as there is no patch for these bugs when they're discovered. Attackers are also able to exploit systems using these vulnerabilities undetected, as safeguards have not been put in place by organizations or enterprises yet.

High-profile zero-day vulnerabilities include Log4Shell, tracked as CVE-2021-44228, a critical RCE bug within Log4j's Java Naming and Directory Interface (JNDI). By exploiting the vulnerability, attackers were able to easily take control of vulnerable systems, a considerable threat as Log4j is used in nearly every Java application.

Other vulnerabilities include PrintNightmare and Proxyshell, both remote execution flaws that were exploited quickly and widely, according to Clark.

"The rise in zero-day exploits is partly driven by just more sophisticated threat actors," Clark said in the Dark Reading webinar. "This can include things like nation-states and also using them in targeted attacks."

Chad Graham, cyber incident response team (CIRT) manager at Critical Start, however, believes that advancements with AI will change the landscape in 2025.

"Both attackers and defenders will rely on AI-driven tools to automate the search for hidden software flaws," Graham says. "This shift will likely result in a more dynamic cybersecurity landscape, where continuous innovation and adaptation become the norm."

Supply Chain Attacks

Supply chain attacks remain an active threat and tend toward the severe as their impact cascades on to multiple parties: customers, suppliers, and other third parties. Attackers exploit a trusted resource and ultimately gain access to not just one organization, but multiple. These kinds of threats remain concerning as organizations depend more and more on outsourcing services.

The best known example is the SolarWinds breach, which impacted the SolarWinds Orion system, at the hands of a group known as Nobelium. More than 30,000 organizations — including state and federal agencies — used the Orion network management system, resulting in the backdoor malware compromising thousands of data, network, and systems.

Tracked as CVE-2020-10148 with a CVSS score of 9.8, the authentication bypass bug allowed an unauthenticated attacker to execute API commands. The attackers in question were advanced persistent threat (APT) actors who infiltrated into the SolarWinds’ supply chain to insert a backdoor.

"The complexity of modern supply chains makes it challenging to secure all the dependencies," Clark said in the webinar. "This underscores the need for rigorous third-party risk management."

In the year ahead, Dana Simberkoff, chief risk, privacy, and information security officer at AvePoint, believes that there will be a sharpened focus on supply chains and third-party risk management.

"The CrowdStrike incident wasn't just a wake-up call — it was a stark reminder that in our interconnected ecosystem, one weak link can trigger a catastrophic chain reaction," Simberkoff says.

Remote Work Infrastructure Exploits

Since 2020 and the COVID-19 pandemic, organizations have leaned into remote and hybrid work offerings, increasing the risk of cybersecurity threats and becoming a significant concern. Attackers focus on vulnerabilities that allow users to engage in remote work such as VPNs, remote desktop protocols (RDPs), and phishing attacks through platforms such as Zoom and Microsoft Teams.

There have been several notable incidents in which VPNs and RDPs were leveraged, allowing threat actors to gain access to enterprise systems and networks. In addition, remote workers are often operating from less secure environments, causing an uptick in phishing attacks as the threat actors try to take advantage of these blind spots.

"The shift to remote work has expanded the overall attack surface, Clark said in the webinar. "Remote workers often need more security controls than those that are working [onsite], which can lead to significant vulnerabilities."

Recent examples of vulnerabilities remote and hybrid work vulnerabilities include CVE-2024-38199, a remote code execution vulnerability (RCE) in the Windows or Line Printer Deamon (LPD) Service, and CVE-2024-21433, a Windows Print Spooler elevation of privilege vulnerability.

"Remote work infrastructure will continue to be a prime target for cybercriminals in 2025, with an increase in sophisticated attacks on cloud services, VPNs, and collaboration tools," says Stephen Kowski, field CTO at SlashNext Email Security+. "We'll likely see more AI-powered threats designed to bypass traditional security measures, exploiting vulnerabilities in interconnected devices and home networks."

Exploitation of AI and Machine Learning Systems

With the rise of AI and its increasing use amid the public, comes widespread risk of exploitation from attackers. Clark noted of adversarial attacks, data poisoning, and model inversion attacks that are at the forefront of emerging threats for AI and machine learning (ML) systems in particular. 

The nature of some ML systems requires feeding a system information for the best results, the system becoming more familiar with the user over time. When attacks target these systems, it can lead to unauthorized access to sensitive data stored and processed within these tools, as well as incorrect predictions or biased decisions.

"AI models will be key areas of exploitation in 2025," says Rom Carmel, co-founder and CEO at Apono. "As AI and machine learning become integral to identity verification systems, attackers will find ways to poison AI models or bypass them."

AI can also simply be manipulated for malicious ends, as seen when an AI deepfake robocall was created to impersonate US President Joe Biden to encourage individuals not to vote in the New Hampshire's Democratic primary, an event that could have had severe consequences on the US electoral process.

"The threat landscape is evolving with the rapid adoption of AI and ML," Clark said in the webinar. "Attackers increasingly focus on these systems to undermine their reliability and exploit vulnerability."

Cloud Misconfigurations

As organizations continue to shift their operations to the cloud, it will continue to emerge as a space for threat actors to thrive, often due to the cloud simply not being set up correctly.

Common examples of threats that circulate within the cloud are publicly accessible S3 buckets, misconfigured security groups in AWS, and exposed databases.

"Cloud misconfigurations can have severe impacts related to data breaches, unauthorized access to critical systems, financial loss, and reputational damage," Clark said. He added that the complexity of these environments is going to increase leading to more frequent configuration errors.

In the past, Amazon and Microsoft cloud environments have exposed customer data, such as viewing habits, names, email addresses, email content, and phone numbers. The leaks aren't due to vulnerabilities but misconfigurations ranging from insecure read-and-write permissions to inaccurate access lists and misconfigured policies.

"To successfully prevent cloud breaches in 2025, companies need to focus on three key areas: visibility, access control, and continuous monitoring," says Jason Soroko, senior fellow at Sectigo. "Cloud environments are dynamic, so your security needs to be dynamic too."

IoT Device Vulnerabilities

IoT devices are allowing for emerging threats to thrive, being easy targets for threat actors to exploit, whether it be due to weak default passwords, lack of encryption, or insecure firmware.

Common attacks that IoT devices face are data theft, network breaches, and distributed denial-of-service (DDoS) attacks. A recent example emerged in the Common Unix Printing System (CUPS) for managing printers and print jobs. The series of vulnerabilities, tracked as CVE-2024-47176, CVE-2024-47076, CVE-2024-47175, and CVE-2024-47177 could allow bad actors to stage DDoS attacks within seconds for less than 1 cent while using an available cloud platform.

"Just the sheer volume of connected devices really exacerbates the threat," Clark noted in the webinar. "Securing these devices becomes really challenging due to their diversity and often limited processing power for adding security features."

And as the use of IoT, OT, and 5G networks continues to rise, organizations will need cyber threat intelligence (CTI) to extend beyond traditional IT environments, says Callie Guenther, senior manager, cyber threat research at Critical Start. "This expansion, which will continue throughout 2025, will add complexity to CTI, requiring more granular insights and specific intelligence data."

Cryptographic Weaknesses

According to Clark, cryptographic weaknesses continue to pose a significant threat because these kinds of vulnerabilities undermine the foundation of secure communication and data protection. These weaknesses often manifest in one of two ways: flaws in encryption algorithms, or how the algorithms are implemented.

"The rising threat is kind of compounded by the fact that as computational capability advances, that previously secure crypto standard now becomes increasingly more vulnerable," Clark said in the webinar.

He recommended regularly updating cryptographic libraries, and enforcing strong encryption protocols to avoid exploitation attempts like man-in-the middle attacks, data integrity issues, and the exposed sensitive information.

Just recently, Acros Security discovered a vulnerability, similar to CVE-2024-38030, that enables an attack in which a vulnerable device is coerced into sending NTLM hashes, which is the cryptographic version of a user's password, to a threat actor.

"We have never before required from [cloud service providers] such granular and detailed information on the type of encryption in use, but customers (government and non-government customers alike) will require this level of detail to ensure their encryption standards are being met," says Philip George, executive technical strategist at InfoSec Global Federal.

API Security Gaps

More organizations are relying on APIs to connect systems; however, these APIs are at risk when they have flaws in the design or the implementation of the APIs. Attackers are able to breach systems through unauthorized access, allowing them to manipulate certain restricted actions.

A notable example of this is the exposure of user data through Facebook’s API, though these flaws are also abundant in other sectors such as healthcare or financial services. 

Gaps in API security ultimately serve as a launchpad, often for data breaches which can lead to the loss of sensitive information, unauthorized transactions, reputational damage, and significant financial loss.

"The threat is escalating as API is becoming more prevalent, increasing the number of potential attack surfaces," Clark said. "To mitigate these risks, it's essential to secure your API endpoints, enforce robust authentication mechanisms, and regularly update and audit API access."

A Docusign API was recently used in a wide-scale phishing campaign due to its "API-friendly environment," which is beneficial for businesses but also provides a way for bad actors to conduct malicious operations. The flaw could ultimately could have led to instances of fraud, though there are ways for users to avoid and detect such API abuse.

In the coming year, the cyber landscape will continue to evolve, API being in the forefront of these changes.

"We anticipate a rise in sophisticated API attacks using automation, artificial intelligence, and advanced evasion techniques to exploit vulnerabilities and bypass traditional security measures," says Eric Schwake, director of cybersecurity strategy at Salt Security. "One significant risk will stem from the exploitation of API misconfigurations, which often occur due to the fast pace of development and deployment. This situation will challenge organizations to adopt a more proactive and comprehensive approach to API security."

Ransomware Evolution

"We could do a whole webinar on ransomware," Clark said in the webinar, which raises the question: Can ransomware even be considered an emerging threat? 

The answer is yes, though ransomware attacks have become one of the most disruptive and costly cyberattacks out there largely due to their rapid evolution.

One of the most notable ransomware attacks occurred on Colonial Pipeline, which shut down its entire operations for the first time, leading to fuel shortages and four states on the East Coast declaring a state of emergency. The ransomware attack prompted action from national security and the executive branch and forced a reevaluation of the nation's critical infrastructure security.

Threat actors know they can win big when demanding ransoms from organizations, such as those in the healthcare sector, which will pay these high prices in order to help patients in need.

"As these attacks are becoming more targeted and, frankly, aggressive, it's crucial to start to implement backup strategies that are robust, strengthen your overall incident response plans, and continuously educate your employees on recognizing and avoiding things like phishing attempts that can often serve as an entry point for ransomware," Clark said in the webinar.

Backups may not always be an option, according to Brandon Williams, chief technology officer at Conversant Group.

"Some threat actors have moved to deleting data as part of their normal motions," he says. "If this gains traction in 2025, organizations will not have a method to recover by simply paying a ransom and hoping to get a working decryption tool. The only method of recovery will be backups; however, data shows that backups do not typically survive these breaches."

5G Network Vulnerabilities

5G networks are being rapidly deployed, and with them come threat actors' awareness and exploitation of its vulnerabilities. Attackers are increasingly able to target 5G infrastructure with ease, and these open the door for even bigger threats such as large-scale DDoS attacks, unauthorized data access, and disruption of our critical services.

"As we consider the rising threat, the global rollout of 5G brings an increasing number of connected devices," Clark said in the webinar. "The rising number amplifies their attack risk, particularly given their reliance on cloud-native infrastructures."

At Black Hat 2024 in Las Vegas, seven Penn State University researchers detailed how mobile devices are at risk of data theft and denial of service due to 5G technology vulnerabilities. Threat actors use these resources simply by providing someone with an Internet connection, allowing easy access to spying, phishing, and more. 

"Vulnerabilities such as lack of initial broadcast message authentication, spectrum slicing, silent downgrade, and unsecured DNS paging currently affect 5G networks," says Mayuresh Dani, manager, security research, at Qualys Threat Research Unit. "In the year to come, these will continue affecting 5G networks and vulnerabilities in unsecured base stations will multiply snooping attacks."