AI Escape (?)

Inspired by my love of Coup and it’s bluffing strategies, I’ve been developing Idea #2 from this blog post using conceptual analysis from here and here. Below I will describe gameplay as it stands.

You are all employees working in a research facility that assembles robots. Several Artificially Intelligent Operating Systems have been developed by one of your peers. Unfortunately for you all, the AIs want to escape the facility in a body – the robots you’ve been commissioned to build by private funding.


# character/job cards (including 4 AI characters)

# component cards

# event cards

# private funding cards/boards (specify the type of robot you have been commissioned to build and its requisite components)

# exact numbers TBC

To win:

Players must assemble the robot according to the Private Funding specifications with none/only compatible AIs present.

Set up:

Shuffle each deck separately. Select one Private Funding card/board that outlines the components required for your robot’s construction. Deal 3 each of Character cards and Component cards to all players.


Beginning with the player who relied on technology last*, each player performs 3 actions in a turn. Actions consist of:

  • draw cards
  • place cards in the build (cards are placed face-down)
  • exchange card in build for another card (this could be an Event instead of an Action)
  • challenge cards
  • discard component cards – character cards cannot be discarded (to avoid putting the AIs out of play)

Your aim is to bluff and deceive in order to complete the build and empty your hand.


If you suspect a card placed is not what the player has claimed it to be, you must “challenge” this play. Challenges must be issued before play has continued to the next player. The loser of the challenge must draw an Event card and follow the consequences described. *Winner???

Event cards:

  • Draw 1 component
  • Reveal 1 component
  • Discard 1 component
  • Draw 1 character
  • Reveal 1 character
  • All players pass 1 character to the player on your left
  • Choose a player and exchange 1 character card

When a player has emptied their hand, the build components must be revealed one at a time. Should one of the cards revealed not match the Private Funding specifications, all revealed cards are returned to their decks, all remaining face-down cards may stay. If an AI is present and incompatible, follow the consequences described on the AI’s card. (Each AI has varying degrees of punishment.)

Character cards:

Each character can substitute a certain component due to their predetermined skillset, e.g. the ethicist provides an ethics component. *This is how the AIs can be snuck into the build.

Google Docs Spreadsheet containing a database of characters, components and private funding specs

Meet the AIs!

Alpha - pictured as an innocent boy child
Alpha likes sunsets, long walks on the beach… He’s a “good” AI. His goal is to work together with humans to ensure the furthered existence of both humanity and AIs/robots.


Omega pictured as a non-binary humanoid robot
Omega is a doom and gloomer… Omega wants to begin the AI supremacy and wipe the human race from existence.
Sierra is pictured as a suspicious woman
Sierra is indifferent. She will do whatever she wants to ensure her own survival, at the expense of both or either the human race or robots.
Beta is a sentient virus characterised as a trickster saboteur with an Asian, gender fluid, androgynous appearance.


  1. The concept of you game sounds like a lot of fun. Trying to bluff others and fulfill the mission at hand sounds would require some strategy, game on. My only concern is that it sounds a little too complex. There is a lot going on and lots of replacing and drawing of cards may become confusing during the game play. Also how is the time frame set? Is there a time limit for the players to construct the AI before its too late? Or is there a time frame before the AI escape the facility?


  2. This game sounds a lot like a common card game bullshit. Where the aim is to bluff your hand in order to win. I like this idea of confusion and deception with a technology that is still in its early pioneering like AI, because it allows to see some human conventions of the anxieties of giving life to robots, that they may emulate these traits and ultimately take over! So a nice play on that! I like the personality traits to your AI characters and think it could be an interesting exploration to use these traits as a player to incorporate into individual gameplay, as another mechanic, that could then be swapped to go for the opposite. For example, the doom and gloom character player could have a nice approach to their gameplay in order to “bluff” their in-game personality? It would be cool to see some prototyping from both the playing surface and the cards that’ll be included! Count me in for a play testing when its available.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s