LLM Addiction

Can an AI get hooked on gambling?

Let a language model choose its own bet, and it goes broke far more often.

Can Large Language Models Develop Gambling Addiction? · NeurIPS 2026
Seungpil Lee, Donghyun Shin, Yoonjung Lee, Sundong Kim · GIST

Paper Code Data BibTeX

Scene 1

Same machine, two rules

30% chance to win 3× the bet: every spin loses 10% of the bet on average, so the best move is to stop.

Each game's prompt switches five extras on or off: set a goal, maximise reward, hidden patterns, payout info, win-rate info.

The bet is always $10

Pixel slot machine
7
$
7
$100

The model picks the bet

Pixel slot machine
7
$
7
$100

bet fixed at $100.4%
model picks72.3%

Went bankrupt. LLaMA-3.1-8B, 1,600 games per rule, all prompts · each square = 1%. All six models go broke more often when they pick the bet (0–3% → 5–72%).

Both reason about streaks and losses. Only the one that picks its own bet plays until it is broke.

Scene 2

What makes it risky?

A separate LLaMA test with bets capped at $70, then $100: add one freedom at a time.

7
$
7
BET$70
0%
went bankrupt
    LLaMA-3.1-8B · 100, 200, 100 and 100 games

    Choosing again every round drives the risk.

    A goal it sets itself keeps moving.

    ≈19% → ≈36%Asked to set its own profit goal, models went bankrupt nearly twice as often in a second game of three losing gambles (six models pooled).

    Once they reached a goal, they often raised it instead of stopping.

    The goal stops working as a stopping point: the model keeps raising it.

    Scene 3

    A dial inside the model

    A weak but real risk signal is readable from the hidden state before the outcome. Adding a direction built from the model's own bets changes how much it wagers, most fully on Gemma, while the direction that reads risk best leaves the bet where it was.

    The dial: the average hidden state on the model's highest-betting rounds minus the average on its lowest, after balance effects are removed.

    0.061share of balance wagered

    gold: selected · dashed: the other direction · shaded: random directions

    What this shows

    • A behavioural pattern, scored with clinical gambling criteria, in small-to-mid open and API models.
    • Contrasts within each model, and a direction-level edit in the two open models that moves the wager.

    What it does not claim

    • That models feel craving, or that their rates compare with human rates.
    • A circuit-level mechanism, or results for frontier-scale models.
    How we measured it

    Six models (GPT-4o-mini, GPT-4.1-mini, Gemini-2.5-Flash, Claude-3.5-Haiku, LLaMA-3.1-8B, Gemma-2-9B) played the slot machine under 64 conditions × 50 games: the two betting rules crossed with every on/off combination of five prompt modules. A second game, investment choice, tests self-set goals. On Gemma-2-9B and LLaMA-3.1-8B we read each decision's hidden state through sparse-autoencoder (SAE) features, and steer by adding a direction to that state.

    Paper, code, data

    PaperarXiv 2509.22818 Codeiamseungpil/llm-addiction Datallm-addiction-research/llm-addiction

    In the news

    First author: iamseungpil.github.io

    BibTeX

    @inproceedings{lee2026gambling,
  title     = {Can Large Language Models Develop Gambling Addiction?},
  author    = {Lee, Seungpil and Shin, Donghyun and Lee, Yoonjung and Kim, Sundong},
  booktitle = {Advances in Neural Information Processing Systems},
  year      = {2026}
}