On April 11, 2026, Vienna became one of the central venues for a global hackathon dedicated to a key question in today’s AI landscape: How can autonomous agents become trustworthy, secure and aligned with governance requirements?
Organised by The AI Impact Mission (AIM) in collaboration with TIMETOACT Group Österreich, Klartext AI, AI Austria, AI Factory Austria AI:AT as location partner, Texterous and Speedinvest, the hackathon brought together a global community of builders. In total, over 800 participants from 88 cities worldwide registered to take part, working both remotely and across more than 20 international hubs.
Photo credit: AIM
Vienna Hub at AI:AT
The on-site event in Vienna took place at the AI Factory Austria AI:AT coworking hub, where around 40 participants worked intensively on the challenge. The focus was on the development of personal autonomous agents that are not only efficient, but also resistant to manipulation and in line with governance requirements.
Participants developed and benchmarked their solutions on the BitGN platform, which provides a structured environment for evaluating agent performance against measurable and transparent criteria. This reflects a broader shift in AI development – away from pure experimentation towards reproducibility, evaluation and real-world applicability.
Programme and collaboration
The hackathon combined hands-on development with continuous feedback and exchange. Teams explored different technical approaches, including:
- Secure and governance-compliant agent architectures
- Evaluation and benchmarking of agent behavior
- Locally runnable models for increased data sovereignty
- Strategies to improve robustness against adversarial inputs
A particularly notable contribution came from a team working entirely with locally deployable models, highlighting the potential for sovereign and secure AI systems.
Results and findings
First place in the Vienna On-Site Challenge went to Bernhard Götzendorfer with 79 out of 104 points. We are particulary delighted about the outcome as Bernhard recently became a member of the AI:AT coworking hub – yet another example of the exceptional talent within the AI:AT community
According to Felix Krause, CTO of Klartext AI, who supported the teams on-site, many of the highest-scoring teams demonstrated strong project management rather than traditional engineering backgrounds. This was also the first AIM hackathon where a non-developer won one of the tracks, underlining an important shift: building effective AI prototypes is becoming increasingly accessible.
Beyond individual results, the hackathon demonstrated that progress in trustworthy AI is driven not only by technical solutions, but by shared learning, comparison of approaches and open exchange.
The role of AI:AT in the ecosystem
For the AI Factory Austria AI:AT, hosting the hackathon at the coworking hub is part of a broader objective: supporting the development of applied and trustworthy AI by connecting people, infrastructure and expertise.
Events like these emphasize the value of collaboration across the entire ecosystem. By cooperating with organizations such as AI Austria, Klartext AI and AIM – AI Impact Mission, AI:AT helps to create spaces where experimentation, knowledge transfer and practical implementation are possible.
As autonomous agents evolve, questions of trust, evaluation and governance remain central. Initiatives such as this hackathon provide a concrete framework to address these – through practical development, measurable benchmarks and international collaboration.
