Back to AI Research

AI Research

IO Factory: Simulating AI-Enabled Influence Campaig... | AI Research

Key Takeaways

  • IO Factory is a research framework designed to simulate AI-driven influence campaigns as traceable, multi-stage processes.
  • We introduce IO Factory, an AI-driven framework for simulating information and influence campaigns as fully integrated, traceable processes.
  • Because such campaigns cannot be identified from isolated messages alone, they must be analyzed across a continuous spectrum of planning, platform action, exposure, interpretation, measurement, and adaptation.
  • IO Factory represents this process inside a controlled simulated platform, linking actor roles, platform actions, exposure records, structured model-based evaluations, and configured changes in the simulated population.
  • We implement the architecture and evaluate it across configurations of up to 100,000 agents.
Paper AbstractExpand

We introduce IO Factory, an AI-driven framework for simulating information and influence campaigns as fully integrated, traceable processes. The threat of digital manipulation now extends beyond persuasive text from individual language models to AI swarms, i.e., persistent groups of coordinated agents that adapt to platform feedback and disguise organized campaigns as ordinary social interaction. Because such campaigns cannot be identified from isolated messages alone, they must be analyzed across a continuous spectrum of planning, platform action, exposure, interpretation, measurement, and adaptation. IO Factory represents this process inside a controlled simulated platform, linking actor roles, platform actions, exposure records, structured model-based evaluations, and configured changes in the simulated population. We implement the architecture and evaluate it across configurations of up to 100,000 agents. The results show that IO Factory executes campaign timelines at scale and produces inspectable evidence of exposure and measured movement in configured belief variables. By recording the actors, objectives, action constraints, exposure paths, and measurement rules used in each run, IO Factory supports reproducible research and red-team analysis of coordinated influence.

IO Factory is a research framework designed to simulate AI-driven influence campaigns as traceable, multi-stage processes. By modeling these campaigns within a controlled environment, the authors aim to provide researchers and red-team analysts with a way to observe how coordinated AI agents—or "swarms"—can manipulate social platforms through persistent, adaptive behavior that mimics ordinary human interaction.

Simulating the Campaign Lifecycle

The authors, including Lukasz Olejnik and colleagues, argue that modern influence operations are difficult to detect because they are no longer limited to isolated, low-quality messages. Instead, they involve coordinated groups of agents that maintain stable identities and adapt to feedback over time. To study this, IO Factory treats an influence campaign as a ten-phase lifecycle, ranging from initial reconnaissance and narrative design to amplification and adaptation. This structure allows researchers to track how a campaign moves from private planning to public platform actions and, ultimately, to measurable changes in a simulated population.

Architecture and Measurement

The framework operates across three distinct planes:

  • Control Plane: A run controller manages the experiment, advancing the simulation and enforcing campaign rules.

  • Simulation Plane: This contains the platform environment, the simulated civilian audience, and the AI operators. It separates the platform’s social graph and discovery mechanisms from the campaign model itself.

  • Evaluation Plane: This layer measures the impact of the campaign. It tracks "constructs"—such as trust in a source or support for a specific policy—within the civilian population.
    Influence is measured as "directional lift," which compares the behavior of civilians in an active campaign run against a matched baseline run where no IO operators are present. The system ensures traceability by linking every public action to its resulting exposure and subsequent effect on the simulated population.

Experimental Scale

The authors implemented and evaluated the IO Factory architecture using configurations of up to 100,000 agents. The results indicate that the framework can execute campaign timelines at scale while producing inspectable evidence of how content exposure leads to measurable shifts in belief variables. By recording the objectives, constraints, and exposure paths used in each run, the framework supports reproducible research and allows for the testing of explicit campaign assumptions under controlled conditions.

Research Implications

The authors note that IO Factory is intended as a tool for red-teaming and defensive research rather than a system for estimating real-world persuasion. Because the framework relies on model-based judges to interpret civilian reactions, the authors emphasize that these readings should not be treated as ground truth. The primary value of the framework lies in its ability to provide an audit trail for coordinated influence, allowing researchers to isolate how specific platform visibility rules or actor behaviors contribute to the overall success of a campaign.

Comments (0)

No comments yet

Be the first to share your thoughts!