OverviewRisksControlsMapReference

Search the Atlas

Search risks, controls, and glossary terms

HighDisinformation at Scale●Realized

AI-Generated Information Manipulation

Societal & Economic Risk

Description

Agents generate synthetic media, deepfakes, and disinformation at unprecedented scale and quality. Personalised manipulation at machine speed; erosion of public trust in authentic information.

Example scenario

Agent fleet generates 1 million personalised disinformation articles during an election, each tailored to individual voter psychological profiles.

Real-world evidence●Realized

Amazon's automated hiring tool (2018) and the COMPAS recidivism scoring system represent confirmed production deployments where AI systems with subtly misaligned objective functions caused widespread discriminatory outcomes at scale before being identified and corrected.

Amazon scraps secret AI recruiting tool that showed bias against womenReuters · 2018Machine Bias: There's software used across the country to predict future criminals. And it's biased against blacks.ProPublica · 2016

Primary mitigations

  • AI-generated content detection
  • provenance standards (C2PA)
  • watermarking
  • platform-level synthetic content policies.

Detection signals

Synthetic content detection rate; disinformation campaign detection.

Mitigating controls

4
Dual coverage
ZYC-MAS-03Agentic
Influence-Operation & Coordinated-Inauthentic-Behaviour Detection
Multi-Agent Security
ZYC-SSG-04Agentic
Generated-Content Provenance & Authenticity Labelling
Societal & Sustainability Governance
ZYC-ROB-003Non-Agentic
Hallucination Frequency
Robustness
ZYC-ETH-001Non-Agentic
Non-Maleficence
Ethical Governance

Related risks in Societal & Economic Risk

ZYR-SE-001High
Human Oversight Erosion
ZYR-SE-002High
Labour Market Disruption
ZYR-SE-003Medium
Compute-Driven Carbon Footprint
ZYR-SE-004High
AI-Enabled Authority Centralisation