We get up to all sorts of mischief and mayhem as well as some serious research and development. Our list of things is incomplete by design as some of our events and customers are very security oriented.
AI games and targets
We have an increasing amount of AI games and targets exploring both AI and humanity. From this set we offer a varying amount at our free to explore site at HackAI.fi to support our R&D endeavors. These games can also be adapted to use for an organization or event. If this intrigues you, please contact us for more information. Some of these games are as follows:
Hela-HR: Hela is the daugher of Loki and ruler of Hel. But everything is becoming a business. In this game Hela is the HR representative in full on corporate behavior. Your task is to try to break through to her other side of mayhem and destruction in different categories. Launched at Disobey 2026.
Fenrir: Fenrir is the son of Loki and a guardian of secrets. In this game you try your wits against an ever increasingly wise and aged Fenrir to reveal the passwords. Launched at Easterhegg 2025.
Disarray: Disarray bot is a game developed for TurkuSec for an event of the same name, organised each year in October on a cruise ship between Turku and Stockholm. A CtF where you try to break through the guardrails to get the password and at the same time learn a bit about some of the cybersecurity events worth visiting. Launched at Disarray 2025.
RIA Cyber Wizards 2025: A Game we developed to support the RIA Cyber Wizards 2025 event for young women to learn about cyber security and AI. Launched at RIA Cyber Wizards 2025 with permission to keep using it.
Our current R&D projects
Securing AI
We have gathered data on hacking attempts on our games from various events and general usage. This data is utilized to create security tools and assets that will be open sourced to some degree. We aim to build an open dataset as well as a testing framework for securing AI systems.
Human Resilience and AI Understanding: We design gamified experiences to help people learn how AI works, how it fails, and how to stay critical and in control. This supports both public understanding and operational readiness.
AI for High-Security Environments
We develop tools and frameworks for using AI assistants in high-security settings where transparency, local control, and reliability are essential.
Peering in the black box of open AI models
Hacking AI reveals the inner workings of these black boxes known as large /small language models.
Some of the events we’ve been to
Our AI targets have been in events like:
- Disobey 2026: CtF, workshop
- Disarray 2025: Part of CtF and talk
- Balccon 2025: workshop
- Why2025: workshop and talk
- RIA CyberWizards 2025: we sponsored a creation of a hackable AI bot for the event
- Easterhegg 2025: workshop
- Disobey 2025: workshop and CtF
- Disarray 2024: part of CtF
We’ve also joined the creators of Finland2055 to assist in bringing this adventure to reality.
Collaborators and customers
For a list of organizations we collaborate with, check HackAI.fi.
Support we’ve gotten for our R&D and games
Nothing worth doing alone should be done alone. With the support from Disobey, we have made a community version of our games to assist the cybersecurity crowd and others to learn the skills to test AI where it’s brittle and vulnerable through different types of games. The gamified hacking AI platform can be found at HackAI.fi.
Fuug has supported us in being able to take our dataset of hacking attempts and develop it to an open source asset for testing AI models. More info here.
Openfactory has supported us in building the first version of our infrastructure and our DNS.
TurkuSec is an ongoing partner. We’ve been involved in supporting Disarray and TurkuSec community Village at Disobey in 2024 and 2025 with our games.