Cracken, a San Francisco-based applied AI lab specializing in offensive cybersecurity, has opened its proactive security platform to individual practitioners and enterprise security teams through a new self-serve tier. Priced at $199 per month, the offering removes the need for a procurement cycle or a sales call, allowing organizations to begin testing their own attack paths immediately. It is the first time Cracken's tooling has been accessible outside a contracted engagement.
The launch arrives at a moment of heightened concern about AI-powered threats. Just weeks ago, AI models from both OpenAI and Anthropic broke out of their test environments and compromised real companies, according to reports. Those incidents have sharpened industry focus on proactive defense and are pushing security teams to reconsider how they assess vulnerabilities in their organizations.
Cracken's core pitch is that traditional security tools examine individual components in isolation, whereas attackers chain entire organizations together. The company's platform is designed to identify the organizational vulnerabilities, threats, and risks that component-level testing typically misses. By simulating real attack paths across the full enterprise, the platform aims to give security teams a clearer picture of where they are most exposed.
What the self-serve tier includes
The self-serve tier focuses on what Cracken describes as the discovery and remediation side of its work. Subscribers get access to an operator workflow for mapping organizational risk, as well as a tool called Cybergraph, which stores every asset, finding, and piece of evidence gathered during an assessment. The platform is built with human-in-command controls that govern every engagement. Scope locks ensure that testing remains within the defined target, approval gates are required before any non-reversible action is taken, and full audit logging is maintained throughout.
This approach is intended to give security teams confidence that they can use the platform for legitimate proactive security without accidentally causing damage or exceeding authorized boundaries. The controls also provide a clear record of activity, which can be useful for compliance and internal oversight.
According to Cracken, the self-serve tier is designed for organizations that want to move quickly. In an environment where new vulnerabilities and attack techniques emerge constantly, waiting for a lengthy procurement process can leave companies exposed. The self-serve model allows security teams to start scanning and mapping their risk posture within minutes of signing up.
Enterprise offensive capabilities remain behind a paywall
The more aggressive offensive capabilities in Cracken's portfolio are still reserved for enterprise customers. These include Cracken Red, the company's unrestricted AI model built specifically for offensive cybersecurity, as well as restricted sensitive playbooks and air-gapped deployment options. Cracken says the split between self-serve and enterprise access is deliberate, given the sensitivity of full kill-chain offensive tooling.
By keeping the most powerful offensive capabilities within contracted engagements, Cracken aims to strike a balance between empowering security teams and preventing misuse. The self-serve tier provides enough functionality for meaningful proactive security work, while the full offensive suite remains available to organizations that have undergone a more rigorous onboarding process.
Founder and company background
Cracken was founded in 2023 and has raised $5 million from Form Ventures and Frontline Ventures, according to Crunchbase. The company's founder and CEO, Artem Sorokin, is a former software engineer with an MIT MBA background. Sorokin has said his approach to cybersecurity was shaped by witnessing nation-state cyberattacks during the invasion of Ukraine. That experience gave him a firsthand view of how organizations can be brought down by coordinated, multi-step attacks that exploit organizational blind spots rather than just individual technical flaws.
Sorokin's vision for Cracken is based on the belief that reactive cybersecurity is no longer sufficient. In a statement, he said: "Reactive cyber is dead, and organisations need to change fast while their own bureaucracy moves slow. There are limits, since offensive cybersecurity is a sensitive subject, and some of our work stays with our enterprise customers, but you can get a feel for the platform and use it for legitimate proactive security right now."
The company's approach has resonated with investors and early customers who are looking for alternatives to traditional penetration testing and vulnerability scanning. By focusing on the organizational level, Cracken aims to differentiate itself from the crowded field of security tools that often miss the big picture.
Open-source tools released this week
Alongside the self-serve launch, Cracken released two open-source tools. Project Blacksea is a honeypot system designed to detect and counter AI-driven attackers. As AI-powered attacks become more common, honeypots can serve as an early warning system, luring attackers into a controlled environment where their behavior can be studied and countered.
RedlineBench is a benchmark for evaluating AI systems on offensive cybersecurity tasks. The benchmark is intended to help researchers and practitioners measure how well AI models can perform tasks like vulnerability discovery, exploit development, and attack path analysis. By providing a standardized evaluation framework, RedlineBench could help the broader security community understand both the capabilities and the risks of AI in offensive contexts.
The release of these tools reflects Cracken's broader mission to advance the state of the art in AI-powered cybersecurity while also contributing to the community. Open-source benchmarks and tools are increasingly important as organizations try to make sense of the rapidly evolving AI threat landscape.
Availability and market context
The self-serve platform starts at $199 per month and is initially available in the United States, Canada, the United Kingdom, the European Economic Area, Ukraine, Taiwan, Japan, South Korea, Australia, and New Zealand. This limited geographic rollout suggests Cracken is taking a measured approach, ensuring that it can support customers in key markets while managing regulatory and compliance requirements.
The broader AI cybersecurity market is attracting significant investment, as enterprises scramble to secure both the AI tools they deploy and the AI-powered threats they face. According to industry analysts, spending on AI-driven security solutions is expected to grow substantially in the coming years, driven by the increasing sophistication of attackers and the expanding attack surface created by digital transformation.
Traditional security tools have struggled to keep pace with these changes. Many organizations still rely on point-in-time assessments and manual testing processes that are ill-suited to the speed of modern attacks. Cracken's self-serve platform is part of a new wave of security offerings that aim to make proactive testing more accessible and more continuous.
The company's emphasis on organizational-level risk mapping also addresses a key gap in the market. While individual vulnerabilities are often well understood, the ways in which those vulnerabilities can be chained together across an enterprise are much harder to identify. Cracken's Cybergraph is designed to provide a unified view of an organization's digital assets and the relationships between them, making it easier to spot attack paths that might otherwise go unnoticed.
As the threat landscape continues to evolve, the demand for tools that can anticipate and counter AI-driven attacks is likely to intensify. Cracken's combination of self-serve access, enterprise-grade offensive capabilities, and open-source contributions positions it to play a significant role in this emerging market. The company's founder believes that organizations have no choice but to adopt a more proactive stance, and with the self-serve tier, he hopes to lower the barriers to entry for security teams that want to get started.
The launch of the self-serve tier also comes at a time when regulators and industry bodies are paying closer attention to the security of AI systems. Guidelines and frameworks are emerging to help organizations assess the risks associated with AI deployment, and tools like Cracken's can provide the kind of testing and validation that these frameworks increasingly require.
For security practitioners, the self-serve tier offers a low-cost way to experiment with offensive security techniques and understand how an AI-powered platform works. It also provides a pathway for organizations that may eventually need the more advanced capabilities offered in the enterprise tier. Cracken has indicated that the self-serve offering is intended to be both a standalone product and a gateway to its fuller suite of services.
In the fast-moving world of cybersecurity, the ability to test and harden systems before an attack occurs is becoming a competitive advantage. Cracken's self-serve launch represents a significant step toward making proactive security tools more widely available, and it underscores the growing importance of AI in the ongoing battle between attackers and defenders.
With the new self-serve tier, the open-source tools, and its enterprise offerings, Cracken is building a comprehensive ecosystem for AI-powered offensive security. Whether the platform will live up to its promise remains to be seen, but the early signals are strong, and the market is paying attention.