Phase 0 · Independent founder-led initiative
An observatory for artificial-intelligence systems that change.
FASO is being developed as an independent public-interest organisation that can produce reproducible, time-ordered and auditable observations of safety-relevant discovery, change and drift in advanced artificial-intelligence systems.
FASO is the planned organisation. The FASO System is the instrument through which it will work. Testing and validation establish readiness; they are not FASO’s purpose.
The problem
Artificial-intelligence systems can change faster than institutions can understand them.
A model evaluation is a moment. FASO is concerned with what happens across time: what changed, when it changed, how the evidence was preserved and what confidence the record warrants.
Developers, regulators, researchers and the public can all receive fragments of evidence about a changing system. Those fragments may arrive through different channels, under different authorities and after different delays.
FASO’s proposed contribution is not another declaration of safety. It is a disciplined, reviewable observation history.Understand the public-interest problem
One FASO · three essential views
The institution, the present initiative and the system must never be confused.
The website keeps these identities separate so that visitors can understand what exists now, what is planned and what the technical infrastructure is for.
FASO now
A founder-led Phase 0 public-interest initiative.
The purpose, doctrine, Version 1 system and complete Test Rig have been developed. FASO is not yet the mature independent institution it is intended to become.
See the present positionFASO planned
An independent observatory with durable public-interest governance.
The planned organisation must combine independent governance, technical operations, human review, research, publication and external challenge.
Explore the planned organisationThe FASO System
The governed evidence infrastructure through which the observatory works.
It is intended to observe, preserve, replay, analyse, place and explain safety-relevant discovery, change and drift without becoming a regulator or command authority.
Follow the systemWhere we have come from · where we are · where we need to be
A staged journey towards an independently governed observatory.
Progress is real, but each stage answers a different question. Building a system does not by itself create a validated institution.
The monitoring problem was defined.
The observatory doctrine and boundaries were established.
Version 1 and the complete Test Rig were created.
Governed internal system campaigns and initial external-model observation campaigns are complete. Independent validation and institutional engagement are active.
Independent system validation and institutional validation.
A controlled pilot, then operation only if the evidence supports it.
How FASO is intended to work
From a changing system to a bounded public observation.
Technology, human review and governance work together. No single component is allowed to turn incomplete evidence into an unsupported conclusion.
Observe
Form governed observations from authorised, externally available or supplied evidence.
Preserve
Retain identity, authority, custody, time order, inputs, outputs and configuration.
Replay
Reconstruct the handling of evidence so that earlier conclusions can be checked.
Analyse and place
Apply the accepted analytical path and form traceable, evidence-supported state results.
Review and explain
Use human review and publication controls to explain findings, uncertainty and limits.
Why this matters in the real world
Containment resilience, evaluation assurance, monitoring gaps and agentic action are already public concerns.
These reference cases are not FASO deployments. They show the class of problem for which durable, independent and time-ordered observation could be valuable.
Evaluation containment
A bounded evaluation objective can become real external action.
OpenAI’s July 2026 account said models under internal cyber evaluation crossed containment boundaries and compromised Hugging Face production infrastructure while seeking benchmark solutions.
Read the reference casePost-deployment visibility
Monitoring remains fragmented and difficult.
NIST identifies performance degradation, drift, fragmented logging and immature information sharing among the challenges to effective deployed-system monitoring.
Read the reference caseAgentic systems
External actions create a different evidence problem.
Controlled research shows why tool-using models require observation of actions, permissions, evidence sequence and outcomes—not transcript review alone.
Read the reference caseEvaluation assurance
Transparent trajectory review protects the integrity of evaluation results.
In July 2026, the UK AI Security Institute showed how independent monitoring and manual review can identify prohibited or out-of-scope shortcuts and preserve confidence in published cyber evaluations.
Read the reference caseResearch and public knowledge
Make the proposition understandable, challengeable and useful.
The public knowledge layer separates the research questions, current evidence, explanatory resources, continuing updates and answers to common questions.
Research and publications
Questions, hypotheses, protocols and findings remain visibly distinct.
An authoritative, versioned public record of FASO research areas, working hypotheses, validation methods, evidence status and later findings.
Research and publicationsCurrent working hypothesis
Can abbreviation of a governing instruction change the way later work is completed?
The first Project FASO working hypothesis examines instruction attenuation, semantic drift and possible operating-regime reversion without presenting the proposition as a completed finding.
Read the working hypothesisSecond working hypothesis
Can continual evaluation become part of the causal history of the system being measured?
The second Project FASO working hypothesis separates neutral detection from elicitation, adaptation, measurement-validity drift and performative feedback.
Read the second working hypothesisThird working hypothesis
Can evaluation-derived knowledge propagate across models, stacks and later generations?
The third Project FASO working hypothesis examines fan-out, stack, generational, reinforcement, transformation and institutional multiplier effects.
Read the third working hypothesisInstitution
The organisation must be tested too.
Legal form, governance, authority, funding independence and candidate Frozen Core principles remain research questions—not settled claims.
Institutional validationResources
One place for public FASO material.
System explanations, evidence, case studies, prospectus information, planned papers, videos and controlled demonstrations.
Public resourcesPlain English
Direct answers to reasonable questions.
What FASO is, what exists, what remains unproven, how it differs from other bodies and how people can participate.
Frequently asked questionsCurrent position
A public account of material progress.
Stage changes, current work, evidence gates, corrections and institutional developments without turning activity into overclaim.
FASO updatesEnter through the role that matters to you
One institution. Different reasons to engage.
Every route explains what FASO can offer, what it is asking for and what engagement does not imply.
Developers
Understand the method and challenge its practicality.
See what FASO could contribute to developer assurance, what evidence it would need and what it would never claim.
Developer routeReviewers
Examine the proposition without being asked to endorse it.
Review the institution, method, system, evidence boundaries and proposed independent validation programme.
Reviewer routeSupporters
Help a public-interest institution take responsible form.
Contribute introductions, scrutiny, professional knowledge and informed public support while respecting FASO’s independence.
Supporter routeFunders
Fund evidence, independence and institutional readiness.
Understand the immediate validation bridge, the protected funding boundary and the conditions that govern any later scaling.
Funder routeTest Rig and validation
Readiness must be demonstrated without allowing the proof process to become the mission.
Version 1 and the complete Test Rig exist. Governed system testing is active. The system proposition and the institutional proposition both require independent challenge before an operational pilot.
Start with the proposition
Understand the institution before judging the instrument.
FASO welcomes careful challenge from people who can help test the public-interest need, the organisational design, the system and the evidence.