Autonomous Replication

Risk

Autonomous Replication

AI systems capable of copying themselves and acquiring resources without human oversight

CategoryAccident Risk

SeverityHigh

Likelihoodmedium

Timeframe2027

MaturityEmerging

Capabilities

7 words · 1 backlinks

This page is a stub. Content needed.

Top Related Pages

Organization

METR

Model Evaluation and Threat Research conducts dangerous capability evaluations for frontier AI models, testing for autonomous replication, cybersec...

Capability

Agentic AI

AI systems that autonomously take actions in the world to accomplish goals, representing a significant capability jump from passive assistance to a...

Risk

Erosion of Human Agency

AI systems erode human agency through algorithmic mediation affecting 4B+ social media users, 42.3% of EU workers under algorithmic management, and...

Approach

Dangerous Capability Evaluations

Systematic testing of AI models for dangerous capabilities including bioweapons assistance, cyberattack potential, autonomous self-replication, and...

Organization

Apollo Research

AI safety organization conducting rigorous empirical evaluations of deception, scheming, and sandbagging in frontier AI models, providing concrete ...

Autonomous Replication