Article
AI Deep Dive Part 2: Data Privacy Concerns
Arete Analysis
Cybersecurity 101

A few weeks ago, Arete’s Threat Intelligence team outlined the history of artificial intelligence (AI). Today, we continue that conversation, exploring data privacy concerns associated with AI tools. AI use cases are often showcased to consumers without warning of potential dangers in their application. When a service is free, your data is often the cost of entry.
Today, we dive into three key elements of data privacy concerns in AI:
What information are you exposing publicly?
What data are you putting into AI applications?
And finally, how are you storing your data?
Operations Security (OPSEC): What information are you exposing publicly?
The public release of information can lead to both positive and negative outcomes. Classification by compilation, in which a series of seemingly harmless pieces of information are pieced together in open source, leading to exposure of proprietary, sensitive information, gives credence to the age-old saying, “Loose lips sink ships.”
You may be wondering what this has to do with AI. Any information posted publicly can be used by developers to train AI algorithms. This could lead organizations to aid their competitors indirectly, should they choose to use the same AI platforms. An example of this is a 2023 lawsuit filed by artists against a number of companies that own AI image-generating tools. The artists argued that the AI companies used their art to train algorithms without the artists being properly compensated. The court ultimately ruled against the artists, demonstrating that it is extremely difficult to prove what data was used to train AI algorithms.

What data are you putting into AI applications?
As the use of AI continues to expand, users should carefully consider what data they are exposing. When using popular public-facing AI platforms, such as those created by OpenAI, Microsoft, and Amazon, users must be aware of the type of data they input. Sensitive data, including client information, PII, and trade secrets, should not be used to prompt public-facing AI tools. Inputs into these tools are used to further train the algorithm and develop these tools.
How are you storing your data?
When an organization decides to create or collaborate on a new AI model, large amounts of data are required to train it. When considering where to store such data, cloud storage appears as an attractive option. However, it is also important to consider the options and risks associated with data storage.
One example of such risk is the May 2024 data breach suffered by cloud-based data storage company Snowflake.

The threat actor responsible for the breach, UNC5537, subsequently extorted Snowflake, leading to at least $2.7 million in ransom payments for data suppression. This attack was primarily driven by compromised credentials without MFA, demonstrating the need for organizations to not only assess their third-party risk exposure but also continually implement security best practices.
Conclusion
AI is a powerful tool for organizations looking to enable employees to work within their strengths and increase efficiency. However, the improper use of AI can have disastrous effects. It is important for organizations to develop policies and training on the implementation and use of AI to set employees up for success and ensure the security of their environments. Tune in next week for the final installment of Arete’s AI Deep Dive: Understanding Biases & How Threat Actors Use AI.
Sources
Back to Blog Posts
Article
Phishing-as-a-Service Evolves with Venom
“Whaling” has taken on a new meaning with a highly targeted phishing campaign active from November 2025 through March 2026, aimed exclusively at senior executives from more than 20 industries. The campaign, dubbed VENOM, is a phishing-as-a-service (PhaaS) platform that combines advanced evasion capabilities with immediate persistence of targeted executives. The initial phish impersonates an internal SharePoint document notification and uses embedded QR codes to convince victims to shift to unmanaged mobile devices to bypass corporate security controls. VENOM aims to establish persistence immediately by either registering a new MFA device or retaining long-lived refresh tokens, allowing threat actors to maintain access even after password resets or other base-level remediation efforts.
What’s Notable and Unique
This campaign is unique in its targeted nature of the PhaaS platform rather than broad, sweeping attempts. The threat actors behind VENOM create convincing phishing emails that impersonate SharePoint activity using the victim’s own domain, company name, and even fabricated email threads. These convincing social engineering tactics, combined with the specific targeting of executives, make this an effective capability for cybercriminals.
VENOM operates as a closed-access system, with full adversarial support, but has no public visibility on the dark web or from security researchers. The service likely operates on an invite-only basis, unlike most PhaaS platforms, which typically seek to have as many paying customers as possible. This, among other items such as the sophisticated evasion techniques, indicates a higher degree of sophistication than most other PhaaS offerings.
Either through MFA enrollment or Microsoft Device Code abuse, the threat actor forces the victim to aid them in establishing persistence early in the attack lifecycle. These tactics result in either valid tokens or an additional MFA login method controlled by the threat actor, meaning typical password resets alone are not effective against this technique. Administrators would be required to explicitly revoke sessions and token grants to mitigate the threat actors’ persistence.
Analyst Comments
Oftentimes, MFA is viewed as a one-stop shop to cybersecurity, but tactics such as this show how threat actors can bypass MFA, or worse, use it to establish persistence. Ultimately, this campaign highlights how modern attacks increasingly abuse legitimate authentication workflows rather than attempting to defeat them outright. Defenses that rely solely on MFA without other security posturing, such as continuous session monitoring, token revocation, and identity logging, can leave organizations vulnerable. As attackers shift toward token theft and device trust abuse, incident response and identity security strategies must evolve accordingly.
Sources
Meet VENOM: The PhaaS Platform That Neutralizes MFA
Article
Threat Actors Continue to Leverage BYOVD Technique
Multiple ransomware operations have recently been observed leveraging the Bring Your Own Vulnerable Driver (BYOVD) technique to disable endpoint security controls prior to ransomware deployment. Notably, the Qilin ransomware group commonly leverages a malicious msimg32.dll file loaded via DLL side-loading, along with vulnerable drivers including rwdrv.sys and hlpdrv.sys, to gain kernel-level access and disable security processes. Similarly, Warlock ransomware has been observed exploiting the vulnerable NSecKrnl.sys driver to bypass security controls. The use of BYOVD has also been observed across ransomware campaigns associated with Akira, INC, Medusa, and other threat actors.
What’s Notable and Unique
The Qilin ransomware group employs a sophisticated multi-stage infection chain, leveraging DLL side-loading (msimg32.dll) to execute malicious payloads directly in memory and evade traditional file-based detection. In DLL side-loading, a threat actor tricks a program into loading a malicious dynamic link library. The malware escalates privileges and uses signed but vulnerable drivers (rwdrv.sys and hlpdrv.sys) to bypass security controls, access system memory, and systematically disable endpoint defenses by terminating security processes and disabling monitoring callbacks at the kernel level.
Akira ransomware operators have also exploited the rwdrv.sys and hlpdrv.sys drivers. Additionally, Arete has observed threat actors leveraging multiple other drivers, including the vulnerable TrueSight.sys, to bypass security controls.
Meanwhile, Warlock ransomware operators disguised malicious activity by renaming rclone.exe to TrendSecurity.exe to appear legitimate. The file functioned as a loader, exploiting the vulnerable NSecKrnl.sys driver to disable security processes, while Group Policy Objects (GPOs) were leveraged to systematically disable security controls across the environment.
Analyst Comments
The BYOVD technique, employed by multiple known ransomware operators, reflects a broader shift toward pre-encryption defense evasion, including suppression of Windows telemetry, removal of monitoring callbacks, and abuse of legitimately signed but vulnerable drivers. This technique enables threat actors to evade detection, maintain persistence for extended periods, and maximize the operational impact of ransomware deployment across compromised environments. Organizations should implement strict driver control policies, such as Microsoft’s Vulnerable Driver Blocklist and application control mechanisms. Additionally, enforcing least privilege access, enabling multi-factor authentication (MFA), maintaining up-to-date patching, and continuously monitoring for anomalous driver and kernel-level activity can further reduce the risk of such attacks.
Sources
Qilin EDR killer infection chain
Web Shells, Tunnels, and Ransomware: Dissecting a Warlock Attack
Article
Ransomware Trends & Data Insights: March 2026
The threat landscape in March had a much more even distribution of threat groups than has been observed since the first half of 2025. Although Akira, Qilin, Play, and INC remained among the most active groups, Arete observed 21 unique ransomware and extortion groups in March, compared to only 15 in February. Akira and Qilin’s activity also declined from the previous month; in February, the two groups were responsible for almost half of all ransomware incidents, but in March they only comprised a little more than a quarter of all activity. Arete also observed activity from several emerging groups in the past month, including BravoX, NightSpire, Payouts King, and Securotrop.

Figure 1. Activity from the top 5 threat groups in March 2026
Analysts at Arete identified several trends behind the threat actors perpetrating cybercrime activities:
In March, threat actors actively exploited FortiGate Next-Generation Firewall appliances as initial access vectors to compromise enterprise networks. The activity involves the exploitation of recently disclosed security vulnerabilities, including CVE-2025-59718, CVE-2025-59719, and CVE-2026-24858, or weak credentials, allowing attackers to gain administrative access, extract configuration files, and obtain service account credentials. Arete also observed Fortinet device exploitation involving various threat groups, with the Qilin ransomware group notably leveraging Fortinet device exploits.
Phishing campaigns leveraging OAuth redirection and a resurgence of Microsoft Teams–based social engineering were also observed in March. In one campaign, attackers sent emails disguised as Microsoft Teams recordings or Microsoft 365 alerts, redirecting victims through legitimate OAuth endpoints to attacker-controlled pages hosting malicious ZIP payloads. A separate campaign has been ongoing since last year, in which attackers flood users’ inboxes with spam and impersonate IT support personnel to trick victims into initiating remote support sessions via tools like Quick Assist.
Arete recently released its 2025 Annual Crimeware Report. Leveraging data and intelligence collected during ransomware and extortion incident response engagements, this report highlights notable trends and shifts in the threat landscape throughout 2025, including Akira’s unusually high activity levels in the second half of 2025, evolving social engineering techniques, and trends in ransom demands and impacted industries.
Sources
Arete Internal
Report
Arete's 2025 Annual Crimeware Report
Harness Arete’s unique data and expertise on extortion and ransomware to inform your response to the evolving threat landscape.



