Recently, a threat actor paired ARTEX, an AI-driven testing framework, with large language models to target South Korean financial organizations, a clear sign that attackers no longer wait for a vulnerability to be published. That is the problem Raven.io was built to solve, and on November 12 we officially entered the Korean market at our first press briefing in Seoul, in partnership with The Executive Action.
A campaign that shows where attacks are heading
Korea is not an isolated case. Around the world, a run of runtime attacks and vulnerability disclosures has put the open-source code inside everyday applications under a spotlight:
- PHP2Shell, Node2Shell and Log4Shell 2.0 exposed critical weaknesses in components that sit inside countless web applications and servers.
- Incidents involving Hugging Face and RubyGems showed how widely shared software ecosystems can become an entry point.
The common thread is simple. The open-source libraries enterprises depend on are public, so attackers can study them just as closely as defenders can. With AI, they can do it faster.
Why waiting for a CVE is no longer enough
Most application security still starts with a published vulnerability. A CVE is announced, a signature is written, a patch is scheduled. But when exploitation can begin before anyone discloses the flaw, that model leaves a dangerous gap.
“Businesses cannot afford to make a published CVE the starting point for protection,” said Roi Abitboul, Co-Founder and CEO of Raven.io. “Attackers can examine the same open-source code that enterprises depend on. Security teams need to understand what that code is doing at runtime and be able to stop exploit behavior, including when no known vulnerability or attack signature exists.”
Protection inside running applications
Raven.io takes a different starting point: how code actually behaves while it runs. Our platform monitors open-source components at runtime, giving security teams visibility into the libraries, functions and execution paths inside their live applications.
That visibility powers runtime prevention. Raven.io is designed to recognize and block abnormal execution without depending solely on published CVE records or known exploit signatures.
Here is a simple example. A library built to decode data should never need to launch an external command. If it suddenly tries to, something is wrong, whether or not a CVE exists for it yet. Catching that unexpected behavior gives teams a chance to stop an attacker who is trying to turn a trusted component into an entry point. It is the kind of protection that matters when AI-assisted attackers are probing for weaknesses nobody has disclosed.
Built for Korea’s regulated enterprises
It is no coincidence that the ARTEX campaign went after financial organizations. Banks, insurers and other regulated businesses hold sensitive data, must keep services running without interruption, and often run complex application environments built up over many years.
Raven.io supports diverse operating systems and programming languages, so it can bring runtime visibility and protection to environments with multiple application stacks and interconnected dependencies. Our partner, The Executive Action, pairs that technology with local market expertise, support and engagement.
“Korean enterprises need protection that accounts for the complexity of their existing application environments,” said Adrian Ong, Managing Director at The Executive Action. “Our role is to work closely with local organizations to understand their requirements and support their evaluation and adoption of Raven’s runtime protection.”
What’s next
At the briefing at the JW Marriott Hotel Seoul, we presented our Korean market strategy and our approach to application protection to local media. Watch the highlights below.
From here, we plan to expand adoption across financial services and other regulated industries, deepen our customer and partner relationships, and extend runtime protection across increasingly diverse application environments.
AI is changing how attacks start. Protection has to change too.
Want to see how Raven.io stops exploits at runtime, even before a CVE exists? Book a demo

