When celebrating its sixth anniversary earlier this year, HarfangLab announced its strategic priorities to deploy more endpoint security solutions and to invest in CTI and artificial intelligence. Since its creation, HarfangLab’s mission has been to provide analysts and cyber professionals with the best tools to empower their mission to protect organizations from all cyber threats, current and future. Today, HarfangLab is proud to introduce Ashley and Kio, its two Artificial Intelligence technologies dedicated to detecting threats and supporting analysts.
With these two complementary engines responding to the different needs of analyst teams, HarfangLab fully leverages artificial intelligence to strengthen endpoint protection.
Ashley, the threat hunter
Ashley, formerly known as the HL-AI engine, is integrated into the agent and built with Machine Learning and Deep Learning algorithms. Ashley analyzes binary files and PowerShell scripts to provide state-of-the-art detection capabilities, even for unknown threats.
The engine is constantly trained thanks to the quality of the data provided by HarfangLab’s threat research teams and partners, enabling sec teams to stay ahead of the curve amid the evolving threat landscape.
Ashley is entirely developed and trained by HarfangLab engineers. It is integrated directly inside the agent, enabling it to detect and protect endpoints even if the agent is disconnected from the console. It is also designed to preserve workstation and server performance.
Ashley is composed of multiple complementary engines:
- Ashley Binaries, which analyzes executable files and libraries to detect threats, including those not yet included in the signature databases.
- Ashley Scripts, which identifies malicious Windows scripts, even obfuscated ones.
Kio, the analyst’s personal assistant
Kio (a.k.a. “Know it Owl”) is the analyst’s assistant based on Generative AI. Kio can be found in HarfangLab console interface, taking on several forms: as a wizard ready to explain the console; as a description bar for defining a precise graphical view with the right filters; and soon, as a threat analysis assistant for obtaining contextual information to help analysts understand and prioritize alerts and actions to be taken in response to a threat.
Various key features are integrated into Kio:
- Kio Documentation, to ask the assistant in natural language and get specific answers about how to use the HarfangLab console.
- Kio Query, to filter through the useful data for cyber analysts to facilitate investigations among security events.
HarfangLab’s solution is fully developed in-house, hosted on its own servers, and supported by dedicated AI models to ensure total data confidentiality. Kio’s model is trained on HarfangLab’s owned servers and based on open-source models.
A vision for and investments in AI since the beginning
Grégoire Germain, CEO and Co-founder of HarfangLab, explains: “Reducing analysis time, finding traces of cyberattacks hidden in complex information systems, identifying abnormal behaviors before they turn into cyberattacks: these are real challenges for cybersecurity professionals that can be addressed by AI. Since the beginning, HarfangLab has been developing artificial intelligence use cases that directly address cyber analysts’ concerns. The launch of Kio and Ashley continues that tradition. AI built by cyber experts for cyber experts – this is our commitment at HarfangLab, and I’m proud to say that we will only accelerate that innovation in 2025.”
Guillaume Ruty, Chief Product Officer at HarfangLab, adds: “Developing AI just to pile onto the AI trend is not our mission. Our AI engine, natively integrated into HarfangLab’s EDR, has existed since 2022. But training it, sharpening it so that it interacts seamlessly with all our other engines and our threat intelligence is the true innovation. Today we are able to provide our customers with highly accurate engines capable of detecting yet unknown threats, as well as provide detailed telemetry, making all these complementary tools highly usable and readable. To put it another way, we’re providing both the haystack and the metal detector to find the needle in it. It’s a perfect application of various AI methods – generative, machine learning, deep learning – to a cybersecurity product, and we’re very proud of it. We are lucky to have top-level AI engineers in our teams, and that’s what sets us apart in the cybersecurity market.”
To learn more about HarfangLab AI, you can visit this dedicated overview: https://harfanglab.io/artificial-intelligence/
About HarfangLab
HarfangLab is a French cybersecurity company specializing in endpoint protection. HarfangLab builds technologies that anticipate and neutralize cyberattacks on devices and servers, while also providing a better understanding of your IT infrastructure for improved security. HarfangLab’s EDR software was the first EDR to be certified by ANSSI, and today protects hundreds of customers worldwide, including administrations, companies, and international organizations operating in highly sensitive sectors. HarfangLab’s solutions are distinguished by their openness, with solutions that integrate natively with all other security bricks; their transparency, as the data collected by the tools remains fully accessible; and the strategic autonomy they offer, as customers are free to choose their hosting mode: cloud, public, private, SecNumCloud, or their own infrastructure.