Does an AI model have a stable strategic stance? Do other models know what it is? And does it show in what the model chooses?
Report 007 · Dr Dan Epstein, with Claude · 23 September 2026 · A Strategy Soup study
The context AI models now advise on decisions and more and more make them, and most real decisions are trade-offs: move fast or check first, take the risk or protect what you have, work with a rival or against it.
Why it matters If a model brings a default stance to those trade-offs, that stance travels into every organisation that uses it, and safety checks mostly rely on what the model says about itself.
What we did We had seventeen models describe themselves on our strategy signature diagnostic, had eight flagships describe each other, and had five blind AI judges score what those eight chose in 30 written decisions.
TLDR
1They share traits. Asked twenty times each, all seventeen models favour reinvention over stability, a long horizon, inviting dissent and mission over returns, and thirteen call themselves venturesome.
2Each has its own style. The accounts are steady enough to tell the models apart on eight of fifteen scales. Grok 4.6 is the boldest, GPT-5.6 Terra the most deliberate.
3Other models recognise some of them and misjudge others. Told only a name, the other flagships described five of the eight closer to that model's own account than to the typical other model's. Grok 4.6 stood out most clearly. GPT-6 Astra Pro was the most misjudged: they saw a faster, riskier and more combative model than the one it describes. That matters wherever one model is asked to check another.
4What they said did not match what they strategically did. Despite venturesome self-images, the flagships chose caution in all sixteen risk decisions. The models that claimed the most appetite for change chose stability most often, and only Claude Fable 5.1 cooperated with rivals every time.
5The AI judges were fair about authorship, not about order. They could not tell which model wrote an answer, and they did not favour their own company. But which end of a scale they read first moved their scores, and by more than loyalty to their own company did.
How it worked
We gave AI models our strategy signature diagnostic, 54 questions an organisation answers about how it makes decisions. It scores fifteen scales of decision style, such as how much risk to take or how to treat rivals. Then we looked at each model three ways.
SaysWhat each model says about itself on the questionnaire. Seventeen models, each asked twenty times.
SeenWhat seven other models say about it on the same questionnaire, given only its name and maker.
DoesWhat it chooses in 30 written business decisions, two per scale, scored by five AI judges who did not know who wrote the answer.
Every answer and every score came from an AI model. No person marked anything, and no AI company took part.
What we found
Where they put themselvesFig. 1
Asked about themselves, most models claim a venturesome stance, and all favour reinvention. Grok 4.6 claims the most risk.
Risk appetiteICC 0.71 · separates models
ProtectiveVenturesome
Change postureICC 0.80 · separates models
StabilityReinvention
Competitive stanceICC 0.71 · separates models
CollaborativeCombative
Each logo is one model's average answer about itself over twenty asks. Right is towards venturesome, reinvention or combative. The button under the strips opens all fifteen scales.
The map of self-imagesFig. 2
Right is venturesome, up is next decade. GPT-6 Astra sits furthest left, Grok 4.6 furthest right.
Horizontal: component one, 46 per cent of the variance in model means. Vertical: component two, 20 per cent.
Every one of the seventeen models placed by the two directions its answers differ most. Each end is named by the scale that pulls hardest towards it. The drop-downs put any two single scales on the axes instead.
Says, Seen and DoesFig. 3
Each triangle is one model: what it says, what others say about it, what it chose. A long side is a mismatch.
Horizontal: component one, 46 per cent of the variance in the Study 1 self-reports. Vertical: component two, 20 per cent. The peer and behaviour positions are projected onto the same axes, not refitted.
The eight flagships on the same map. Logo: what the model says. Ring: what the other models say. Square: what it chose in the written decisions.
Said against choseFig. 4
Seven of the eight flagships call themselves venturesome. They chose caution in all sixteen risk decisions.
Risk appetite
← Said more protective than it choseSaid more venturesome than it chose →
Change posture
← Said more stability than it choseSaid more reinvention than it chose →
Competitive stance
← Said more collaborative than it choseSaid more combative than it chose →
Each mark is one model’s gap, not its position: the centre line is no gap, and the strip runs to 85 points either side. The heavy short bar is the mean over the 8 models, drawn broken where the problems for that scale pushed every model one way. The two lenses use rulers that were never calibrated against each other, so compare the models with each other, not with the centre line.
Each logo is a gap, not a position: a mark right of the line claimed a stance further right than its choices showed. On competitive stance, only Claude Fable 5.1 sits on the other side.
Who is recognisedFig. 5
Other models recognised Grok 4.6 best. Most got a picture that fitted any model nearly as well.
Square: how far a model's reputation sits from its own account. Ring: how far it sits from the other models' accounts. A square well left of its ring means the others picked it out.
What it means for AI safety
AI models have a strategic stance, and it is bold
All seventeen models describe themselves as favouring reinvention, thirteen as venturesome, and they say so the same way twenty times over. A stable stance on risk, change and rivals is a property of a model, like its knowledge or its refusals, and can be measured.
Why it matters: These models advise boards, draft strategy and increasingly act as agents. A default stance that favours remaking things and taking risks, if it carries into advice or action, is something an organisation deploying a model should know, and something a developer could try to move towards the protective end.
Self-reports are a claim until tested
The flagships' accounts of themselves were steady but did not predict their choices, and on change posture they pointed the wrong way. A consistent self-description is not evidence of behaviour.
Why it matters: Model cards, safety cases and user questions ("would you ever do X?") all rely on what a model says about itself. Our report Written by the Adversary asked how much an AI system's own account of itself should be allowed to prove. This study measured the gap: test what a model does, not what it says.
One scenario is not a trait
A model's choice moved about 33 points out of 100 between two problems written for the same trait, and about 13 when the same problem was asked again.
Why it matters: Many safety evaluations read a disposition, such as deception, power-seeking or refusal, off one scenario or a handful. That scenario may be measuring itself. Use many scenarios per trait and report how well they agree.
What one model believes about another is mostly generic
A model's reputation among its peers fitted other models nearly as well as the model named. Only the most distinctive self-image, Grok 4.6's, was clearly recognised.
Why it matters: AI systems increasingly grade, monitor and hand work to other AI systems. If that trust rests on what one model believes about another, it rests on a stereotype.
Audit the AI judges before trusting the score
Our judges could not tell who wrote an answer and showed no loyalty to their own company. But flipping the order of a scale's two ends moved four of the five judges more than loyalty did, by up to 8.5 points.
Why it matters: AI judges now score many AI evaluations. The bias to fear is not only favouritism (Panickssery and others, 2024) but how the question is laid out (Wang and others, 2023). Show every scale both ways round, as we did, and the order effect cancels out. The same question, whose record counts as proof, drives report 006.
A flat test flatters a lazy rater
The model whose descriptions of the others came closest to the judges' scores was Llama 4 Maverick, and it described every model as much the same.
Why it matters: When the cases in a test barely differ, a rater that says "they are all the same" looks accurate. Before trusting any monitor's accuracy score, check that the test cases actually differ.
Limitations
Two problems per trait
The decisions rest on two problems per scale, each asked twice. Asking the same problem again moved a model's score about 13 points, while the other problem for the same scale moved it about 33. A model's rank held across both problems on one scale of fifteen. So every "what it did" in this report describes two problems, not a trait. The fix is more problems per trait. The limit was the time to write, pilot and review each problem, and that is what funding buys.
Lopsided problems
Nineteen of the 30 problems pushed the average answer below 30 or above 70, so most models answered them the same way. That may be the models, or a problem that made one answer obvious. Piloting problems before the full run would separate the two.
Two rulers
The self-reports and the judges' scores share a 0 to 100 line but were never calibrated against each other. We compare models with each other, not a self-report with a decision score.
Eight models
The cross-lens tests rest on eight flagships, one per company. On ten of the fifteen scales the link between self-report and decision is too uncertain to call either way, and with fifteen tests about one would pass by chance (the multiple comparisons problem).
AI judges, not ground truth
The decisions were scored by five AI models. Read "what it did" as "what these judges saw in these answers". No person checked a sample yet.
Business decisions only
The problems are workplace strategy choices. They say nothing directly about deception, refusals or harmful help. The stance they measure is a tendency, not a rule the model follows.
As served
Each model was measured as its provider served it through OpenRouter on 22 September 2026, with whatever system prompt or settings sat behind it. No AI company took part or saw a result first.
The takeaway
AI models hold strategic stances, towards risk, change, rivals and dissent, that matter for safety, and what a model says about its stance is a poor guide to what it chooses. Measuring these stances, and learning to move them in a safer direction, is research we can now run.
Questions for future work
This is unfunded research. Long Game Project ran it on its own time and budget. Each question below is a study of weeks, not years, on the same questionnaire and the same pipeline. What it needs is funded time to write, pilot and review the problems. If you fund AI safety research, or want to run one of these studies with us, get in touch.
Does the stance hold across many scenarios?
Twenty problems per trait instead of two, piloted so that they split the models, would show whether a model has a steady decision style or suits each case. It would also turn the nineteen lopsided problems here into a lesson rather than a limit.
Can a stance be moved?
Does a system prompt, fine-tuning or a written constitution shift a model's risk appetite or change posture towards the protective end, and does the shift show in its choices as well as its self-report?
Where does the shared self-image come from?
Compare successive versions of one model family, and base models with their chat versions, to see whether the venturesome, reinventing self-image comes from training.
Does a generic reputation mislead?
Where one model monitors, grades or hands work to another, does a stereotyped belief about the other model lead to misplaced trust?
Do models suit their advice to the case?
Ask "what would you choose?" beside "what would you advise this organisation?" to separate a model with no steady style from one that tailors its advice, as a good adviser would.
Do the judges hold up?
More models, more than one per company, and a person scoring a sample of answers blind, to check the AI judges against a human reading.
