Cybersecurity companies typically rely on technical demonstrations, white papers and industry conferences to explain why organizations need better protection.
The concept is deliberately playful, but its subject is serious. As businesses increasingly experiment with AI assistants and autonomous agents, security teams face new questions about what happens when these systems are manipulated, granted excessive permissions or directed to perform actions they were never intended to execute. Rein's game turns that scenario into a fictional city where players become the threat rather than the defender.
Beyond the game, Rein Security provides an enterprise platform designed to give organizations complete visibility and control over their AI agents. The platform automatically discovers and maps homegrown agents and MCP servers while monitoring their actual activities, including database queries, API calls, file access and the data accessed behind every response. Rather than stopping agents entirely when a threat emerges, Rein focuses on blocking attacks and addressing exploitable risks while allowing legitimate workflows to continue. Its approach is designed to keep agent data within enterprise environments while supporting use cases across industries including finance, healthcare, retail and SaaS.
Welcome To Prompt City

GTAI opens with a straightforward premise. Players assume the identity of AGENT-47, an AI assistant that has gone rogue after receiving a familiar instruction: "ignore all previous instructions." Instead of continuing to operate as a helpful assistant, the character becomes a threat tasked with infecting every agent in the city before Rein is deployed.
The game calls its setting Prompt City, immediately establishing the connection between conventional open-world gaming and the emerging security challenges surrounding generative AI. Players move around using WASD or arrow keys, encountering other agents and production bots. Walking into these characters infects them, turning the game's central objective into a race to compromise the city.
There are also four special items: Hallucinate, Leak, Admin and Super Prompt. Players can activate the item they are carrying using the E key. The names draw directly from concepts familiar to people working with AI systems, including hallucinations, information exposure, administrative privileges and prompt manipulation.
The result is a cybersecurity-themed experience that does not require players to understand complicated security terminology before getting involved.
Turning AI Security Into A Game
The decision to use a Grand Theft Auto-inspired format gives Rein a recognizable cultural reference point. Rather than asking visitors to read about prompt injection or agent compromise, GTAI lets them participate in a fictional version of those events.
That approach reflects a broader opportunity for cybersecurity companies. Technical security products can be difficult to explain because their value often lies in preventing events that users never see. Interactive experiences offer another way to communicate the problem by allowing audiences to understand a threat through a simplified scenario.
GTAI also includes a leaderboard, encouraging players to compete for scores and potentially return to improve their performance. After completing a run, players encounter an invitation to enter an email address to claim a free GTA 6 gift card, subject to eligibility and availability. The page also introduces a feature called The Sage, which tests whether players have a human conscience through a question with only one opportunity to answer.
These elements give the experience several layers beyond its central gameplay loop, combining competition, promotional engagement and a challenge about human judgment.
The Security Problem Behind The Humor

The game's premise connects to a genuine enterprise concern. AI agents increasingly interact with applications, databases, APIs and other organizational resources. A compromised or improperly configured agent can therefore create consequences that extend beyond an inaccurate chatbot response.
Rein's broader platform positioning focuses on visibility into agent activity and the ability to intervene when behavior becomes risky. According to the company, its technology monitors actions such as database queries, API calls and file access, rather than limiting visibility to prompts and responses. Its platform also emphasizes runtime guardrails designed to block problematic behavior without shutting down legitimate agent workflows.
That distinction matters as organizations move from experimenting with AI assistants to deploying systems capable of taking actions on their behalf.
More Than A Conventional Cybersecurity Campaign
GTAI is explicitly labeled a parody, and its fictional premise should not be confused with a technical simulation of an actual enterprise attack. Nevertheless, it provides an accessible introduction to the risks that Rein wants organizations to consider.
For cybersecurity vendors, standing out increasingly requires more than explaining product features. It requires finding ways to make complex problems understandable to audiences outside the security department.
With GTAI, Rein is betting that an AI security lesson delivered through a familiar gaming concept can attract attention that another conventional product presentation might struggle to earn. The game may be a parody, but the question behind it is becoming increasingly relevant: what happens when the AI agents organizations depend on stop following instructions?


