The sixth Human-aligned AI Summer School will be held in Prague from 13th
to 16th July 2026.
Four intensive days of talks, workshops, and discussions covering current
questions in AI alignment research and the broader framings that shape them.
Format of the school
The school is focused on teaching and exploring approaches and frameworks,
less on presentation of the latest research results. The content of the school
is mostly technical – it is assumed the attendees understand current ML
approaches and some of the underlying theoretical frameworks.
Some topics of 2026:
Arguments for AI risk revisited: We will revisit the classical
arguments for AI risk and importance of alignment: risks from
discontinuities, mesa-optimization, scheming – examining how each has
fared against a decade of evidence, how the existing solutions look and how
they may scale.
Post-training and alignment strategies today: We will examine what
frontier AI post-training actually consists of today: character training,
mid-training, and RLVR.
Understanding AIs: How we can understand current AIs using different
approaches and levels of abstraction ranging from mechanistic
interpretability to LLM psychology.
Representational alignment: We will examine work on convergence of
internal representations across capable systems – including the
Natural Abstraction, Platonic Representation Hypothesis and related lines of
inquiry.
Complex systems alignment: AI systems do not exist in isolation, but
are part of a more complex ecosystem of both human and artificial agents and
institutions. What does the alignment relation mean in this context? How do
we align the post-AGI civilization?
The program is subject to change.
The intended audience of the school are researchers interested in learning
more about the AI alignment topics, PhD students, researchers working in ML/AI
outside academia, and talented students. It is recommended that the
participants have a basic general understanding of
the AI alignment problem, although the school does not assume deep knowledge of the topics.
The school will consist of lectures and topical series, focused smaller-group
workshops and discussions, expert panels, and opportunities for networking,
project brainstorming and informal discussions.
Speakers
Ondřej Bajgar – AI Center, Czech Technical University. Ondřej works at
the intersection of technical value learning and AI governance,
investigating how to align AI with human preferences while ensuring robust
safety through principled constraints. His research includes developing more
efficient methods for Bayesian inverse reinforcement learning (IRL) and
other applications of quantified uncertainty in AI alignment, and proposing
novel frameworks for AI regulation, most notably the use of negative human
rights as a foundation for both international policy and technical safety
specifications.
Raymond Douglas – ACS Research. Raymond works at ACS
Research, investigating how increasingly powerful AI systems will fit into
society, and how to make that go well.
Adrià Garriga-Alonso – FAR AI. Adrià works on
mechanistic interpretability and AI alignment, trying to understand what is
happening inside neural networks and what agent-like AI systems actually
want. As a Research Scientist at FAR AI he led interpretability research;
his work on Automated Circuit Discovery (ACDC) helped establish
circuit-based interpretability as a field, and the method is now used across
major AI labs to understand how networks implement algorithms. He also
developed Causal Scrubbing, a rigorous framework for testing
interpretability hypotheses. He is currently building Dokimasia, tools to
help people realize their values when using computers.
Tomáš Gavenčiak –
ACS Research.
Tomáš works on the strategic risks emerging from complex multi-agent AI
systems, investigating how large-scale interactions between autonomous
agents can lead to systemic failures. He applies methods from game theory,
active inference and complex systems modeling to analyze the dynamics of AI
ecosystems. He is also interested in building tools for human cognition and
cyborgism.
David Hyland – University of Oxford. David studies how
bounded rationality shapes coordination in multi-agent systems made up of
both humans and AI. He draws on game theory, collective intelligence, and
models of human cognition to understand why real, resource-limited agents
behave the way they do – and why those limits matter for aligning AI
systems with human interests.
Vojta Kovařík – Czech Technical University. Vojta works
on the theoretical and game-theoretic foundations of AI safety,
investigating how to design robust evaluation and oversight schemes, prevent
systems from misbehaving due to issues like Goodhart's law, issues of
multi-agent alignment, and more.
Jan Kulveit –
ACS Research.
Jan works on the systemic and macro-strategic risks from AI, investigating
how the integration of AI into society could go wrong. His recent research
focuses on the risk of "gradual disempowerment," where humanity could lose
control not through a sudden takeover, but by ceding authority to AI-driven
economic and cultural systems that become misaligned with our core values,
and on "LLM psychology," investigating how we can understand and model the
high-level behavior of large language models.
Ouro Hölzer – ORXL / AFFINE. Ouro works on formal outer-alignment targets with Orthogonal, designing frameworks in which increasingly accurate guesses regarding a formula's solution become benign at the highest degree of optimization. They are currently organizing the AFFINE Seminar, introducing newcomers to theoretical alignment work.
Daniel Polani – University of Hertfordshire. Daniel is Professor of Artificial Intelligence at the
Department of Computer Science, Director of the Centre for Artificial
Intelligence and Robotics Research (CAIRR) and Head of the Adaptive Systems
Research Group at the University of Hertfordshire.
His research interests concentrate on the understanding and modeling
of intelligent decision-making from the bottom up, in the context of
evolution and cognition in artificial and biological agents. His
research includes the emergence of structure in cognitive problems and
the design of biologically plausible decision and control strategies,
mostly through the lens of information theory.
Lily Stelling – Oxford Martin AI Governance Initiative.
Co-author of Europe 2031.
Jesse Hoogland – Resolution.
More speakers will be added soon – please check back for updates.
You can also get a sense of the school by watching talks from HAAISS 2025:
Applications for HAAISS 2026 are open.Apply here. Applications are
evaluated on a rolling basis. We expect most of the spots to be filled by the second week of June.
The price of the school is EUR 300 with a discounted price of EUR 150 for
students, and includes full catering and
conference events.
We have a limited number of scholarships for participants from disadvantaged
backgrounds and in difficult situations, covering the price of the school and
partially subsidizing travel and accommodation costs. If you require financial
support to attend the school, please indicate this in your application and we
will contact you with an application for support.
Proposing workshop sessions
The school will have several smaller-group workshop sessions, consisting of
presenting new topics, interactive activities, and discussion on narrower or
technical topics, offered by both invited speakers and selected participants
of the school.
If you are interested in presenting a workshop, please send us a brief
description of the workshop content (about half a page) along with the
intended format and target audience, noting your expertise within the area.
Note that the expectation is not to necessarily present your own work, but
rather to provide content and activities that enrich the school for other
participants. If we think your workshop is a good fit we will get back to you
and discuss the content of the workshop in more detail and help you refine it
for the school.
Preliminary Program
The detailed schedule will be announced closer to the event – please
check the website for updates.
Full day of lectures, workshops, and discussions, concluding with closing
session around 16:00.
Friday, July 17
No official program. Self-coordination for any unofficial post-school program,
e.g. participant-led sessions and coworking, as well as excursions.
Venues and catering
The school is held at
Břehová 78/7, Praha 1
(Faculty of Nuclear Sciences & Physical Engineering, Czech Technical
University).
Catered lunch, coffee breaks, and light breakfast are provided during the
conference days. Vegan options available. All the school events are covered by
the registration fee.
Organizers
Program:
Tomáš Gavenčiak, Jan Kulveit, Vojtěch Kovařík, and Ondřej Bajgar
Project management:
Anna Gavenčiak
Operations:
Viktorie Havlíčková and Veronika Hendrychová
Organized by
ACS Research
and
Center for Theoretical Study, Charles University, Prague