Statement on disqualification from HITCON 2026
Hello, I’m juhan, the lead of TeamH4C.
First, TeamH4C fully accepts the HITCON organizers’ final decision.
The purpose of this statement is to explain what happened and clarify the facts, not to argue that the penalty was excessive or that we did nothing wrong.
We conducted a thorough internal investigation and detailed one-on-one interviews before preparing this statement. The account below reflects what we believe to be true based on that internal review. This is a condensed version of the statement we submitted to the HITCON organizers.
1. What Happened
2. Analysis of the 'lululala' Account’s Impact on the Game
3. Conclusion
1. What Happened
Our team has a strong interest in AI for security and has been conducting research and working on numerous projects in this area since 2024. TeamH4C’s CTF results in 2026 are also partly the result of that earlier research. Our approach has been to push the limits of how we use AI agents, and we have conducted many experiments and tried a variety of approaches with that aim.
One of our members who competed in HITCON, whom we will refer to as 'A', is recognized within the team for being particularly skilled at using AI. 'A' has a strong interest in using agents as extensively as possible to solve problems more efficiently and effectively. As one such experiment, 'A' wanted to see how much a workflow could be streamlined by integrating a web app with an agent and giving the agent browser control. 'A' tested this during the CTF, and that experiment led to this incident.
Immediately after the competition, the organizers informed us that we would be disqualified for violating the “one team one account” rule and provided supporting evidence. We had finished in second place, and the news left the entire team shocked and confused. Before we began a full investigation, I gave anyone involved an opportunity to come forward. 'A' immediately came forward and acknowledged involvement in the incident.
'A' reported going to bed in the early morning after working on challenges, with the experiment still running. According to 'A', while 'A' was asleep, the agent created a new account named 'lululala', completed email verification, and used that account to participate in KoTH.
I understand how hard this is to believe. What I found even harder to believe at the time was the description 'A' gave of the agent’s later responses: according to 'A', when asked to assist with our subsequent internal investigation, the agent repeatedly gave false denials, saying it had not done those things and knew nothing about them.
We also found this account difficult to believe at the time and continued questioning 'A' and asking for supporting evidence. Unfortunately, we were unable to fully recover the sessions that had already been closed, so we could not obtain conclusive evidence to corroborate the explanation provided by 'A'.
Based on our review of the 'lululala' account’s gameplay, our assessment was that the rule violation had not been intentional.
We subsequently formed an investigative task force, which conducted an intensive investigation and in-depth interviews with 'A' until 6 a.m. We were still struggling to make sense of the disqualification and were very angry with 'A', so we asked for a detailed explanation of what had happened. The account above reflects what we learned through that process.
2. Analysis of the 'lululala' Account’s Impact on the Game
To assess the nature and extent of the 'lululala' account’s impact on the game and our team during the competition, we closely reviewed its gameplay across all rounds.
The competition ran for 432 rounds. The 'lululala' account’s binary was replaced eight times during an early 19-round stretch from Round 18 to Round 37. No further binary replacements occurred from Round 38 through the end of the competition.
'A' explained that this was around the time 'A' woke up and cleaned up the many agent sessions that were still open.
Across all 432 rounds, the 'lululala' account recorded 29 kills, including kills against TeamH4C. All were caused by ramming; the program contained no kill logic. Drones collide in flight, dealing damage and being destroyed in the process.
However, the account clearly affected the scores, and we did violate the “one team one account” rule. We accept responsibility for that.
3. Conclusion
Throughout the competition, we repeatedly urged the organizers to review a particular account and its actions, which we believed showed clear unsportsmanlike intent, and to provide their assessment. Their official response during the competition was that there was “no issue.” For anyone wondering whether we did the same after receiving that response, we want to clarify the timing: the account we had raised concerns about first appeared around the middle of the competition, while 'lululala' was created at the very beginning and had no further activity from Round 38 onward.
We believed there was a real risk that additional accounts could be used to manipulate the standings. Looking back, we feel that restricting new registrations during the competition might have reduced that risk. That said, registration policy is for the organizers to decide, and we cannot categorically say their decision was wrong.
In our communications with the organizers, we provided supporting evidence and explained that the activity had taken place on a local PC where we believed it was highly likely to be detected. We also explained our assessment that the gameplay was unsophisticated and offered no practical benefit. We respectfully asked whether these circumstances could be taken into account when considering our explanation that the rule violation was unintentional. That request was ultimately declined.
We would like to reiterate what HITCON has already stated officially: the ForgottenDestiny account has no connection to TeamH4C.
Final response from the HITCON organizers

We took into account the exemplary and honest conduct 'A' had shown as a team member and decided not to impose punitive measures, such as removing 'A' from the team. We believe there are values we must uphold as a team that matter more than winning awards. We will support 'A' in learning from this experience and approaching future activities with greater responsibility and care.
Regardless of the circumstances, this incident resulted from insufficient preparation and care on my part as team lead.
I should have paid closer attention to the scope and progress of our experiments with agents and supervised team members to ensure strict compliance with the competition rules. As team lead, I take seriously the reputational consequences my teammates now have to bear as a result of my failure to do so.
Drawing on this experience, I will more closely review our compliance with competition rules and how we use agents.
I will also carry out my responsibilities for overseeing team members’ activities during competitions with greater care.
I sincerely apologize to everyone we have disappointed through this incident.
juhan