Artificial Societies

Overview

Artificial Societies is a United Kingdom applied computational social science company that develops AI systems for simulating human audiences and societies. Its enterprise software is intended to help organizations test strategies, messaging, policies, and other consequential decisions against synthetic personas before acting in the real world.

The company positions itself in the emerging market for AI-generated audience research and social simulation. Its notable characteristics are its founders’ behavioral-science and computational-social-science backgrounds, its research involving large agent swarms, and its use of configurable synthetic societies rather than a general-purpose consumer chatbot.

History

Artificial Societies was founded in October 2024 by James He, Patrick Sharpe, and Tom Whittle. He previously researched computational social science at Cambridge and led a peer-reviewed study involving a simulated society of 33,000 agents; Sharpe worked in applied behavioral science and consulting; Whittle has a background in AI safety, public policy, and engineering. The company joined Y Combinator’s Winter 2025 batch and has developed an enterprise simulation platform for testing strategies, messaging, and policies with synthetic audiences. Artificial Societies Ltd is a wholly owned subsidiary of Delaware-registered Societies, Inc.

Product Lines

Manufacturing

Artificial Societies does not manufacture physical products. It develops and operates enterprise software and AI simulation systems, with the platform delivered online; the terms describe standard and bespoke synthetic societies containing AI-generated personas.

Notable Products

• Artificial Societies Platform - Enterprise platform for reviewing and interacting with simulations of synthetic audiences.
• Standard Societies - Prebuilt synthetic audiences generated from audience characteristics or customer-provided social-media profiles, comprising 300 to 2,000 AI personas.
• Bespoke Societies - Custom synthetic audiences built to a customer’s specifications for specific research or decision-making needs.

Reputation

Artificial Societies is presented by technology and research sources as an early-stage company working at the intersection of behavioral science, AI agents, and enterprise research. Its strengths include founders with relevant academic and applied backgrounds, a published large-scale agent-simulation study, and a focused enterprise use case. Public information is limited because the company is young, and independent evidence about simulation accuracy, reliability, pricing, and customer outcomes remains limited. The concept also carries the broader criticism associated with synthetic-persona research: simulated responses may not reliably represent real human behavior without careful validation.