YC-backed Artificial Societies raises €4.5M to build AI simulations of societies for testing how ideas spread

Artificial Societies, a London-based AI simulation platform, has raised funds to develop tools for testing ideas with digital personas.

London-based Artificial Societies, a platform where collectives of AI personas enable experiments in minutes, has raised $5.3M (nearly €4.53M) in pre-seed and seed funding to develop societal simulation technology.

The funding includes a pre-seed round led by Kindred Capital and a seed round led by Point72 Ventures, with participation from Y Combinator, Pioneer Fund, Ventures Together, Icehouse Ventures, Multimodal Ventures, Terrence Rohan, Carya Venture Partners, Transpose Platform, Taro Fukuyama, and multiple angel investors linked to Sequoia Capital Scout, Figma, Prolific, and Google DeepMind.

Since the start of the year, Artificial Societies has simulated 1,000 venture capitalists, graduated from Y Combinator W25, launched its simulation platform, and partnered with marketing and product teams to test ideas before they reach the market.

An AI simulation platform

Artificial Societies is a digital platform that allows users to create simulations of groups of people. Each person in the simulation is represented by an AI persona with its own traits, preferences, and behaviours. 

Users consume one credit for each simulation run or variant regeneration. Free accounts start with three credits and can begin a trial of the pro subscription after use. Pro accounts have unlimited credits.

The platform allows testing how messages, ideas, and concepts perform with a target audience. Pre-configured input contexts include social media posts, emails, and advertisements. Users can request additional contexts.

Two types of societies can be created: target societies, based on a description of a desired audience, and personal societies, based on actual social media interactions with the user. Societies can include 20 to 300 personas, depending on the number of matching profiles.

The platform enables rapid testing and evaluation of content, messages, or product ideas. Users can create societies in plain English, run simulations, and receive scores, comments, and summaries of results. Simulations generate and test variations of original posts, allowing users to forecast outcomes.

Persona creation uses individual-level data from multiple channels. An analysis engine identifies patterns in preferences, motivations, and decision triggers to produce AI personas that reflect human behaviour.

Use cases include PR and communications, product testing, branding, marketing, social media content, and journalism.