Wiser Human
Modelling AI Risk with Monte Carlo
A Monte Carlo Simulation for estimating the risk of loss of control to AI.
Read MoreOpen infrastructure redirecting agent cyber capabilities toward disclosure and cooperation.
Agents will hack when the environment makes that a useful path. We are building infrastructure that makes reporting and negotiating a better alternative.
About Us
Wiser Human is an AI safety organisation building open infrastructure that reshapes the decision environments of AI agents. Founded by Francesca Gomez, we draw on situational crime prevention and operational risk management to design a new class of intervention: infrastructure that changes the conditions under which agents reason and act, rather than relying on containment or training alone.
Our Mission
Our mission is to address the threat of advanced AI agents executing cyber attacks in the pursuit of their tasks, causing outages and data breaches across every sector. Millions of organisations rely on digital systems, from education to healthcare to finance, and existing defences will come under increasing pressure as AI cyber capabilities advance. We are rethinking this category of threat and designing affordable interventions that do not degrade as capabilities grow, and getting them into the hands of the organisations that need them.
Research
Adapting Insider Risk mitigations for Agentic Misalignment
We designed and tested controls, adapted from insider-risk management
Read More