HighPower Concentration○Theoretical

AI-Enabled Authority Centralisation

Description

AI agents enable unprecedented concentration of economic, political, and informational power. Erosion of checks and balances through AI-mediated control; systemic societal risk from power imbalance.

Example scenario

Single organisation controls 80% of economic decision-making through a fleet of autonomous trading and resource-allocation agents.

Real-world evidence○Theoretical

While emergent capabilities in large models are well-documented in research (e.g., Wei et al., 2022), adversarial exploitation of such capabilities specifically before developers identify them has not been confirmed as a production incident. The exploitation pathway remains plausible but lacks empirical cases distinct from general jailbreaking.

No public incident on record — evidence level: Theoretical

Primary mitigations

Diversity requirements in AI deployment
antitrust monitoring
governance diversity standards
concentration metrics.

Detection signals

Power concentration indices; market share metrics; governance diversity metrics.

Mitigating controls

Dual coverage

ZYC-SSG-03Agentic

Authority-Concentration Review & Decentralisation Safeguards

Societal & Sustainability Governance

ZYC-ETH-001Non-Agentic

Non-Maleficence

Ethical Governance

ZYC-ROB-003Non-Agentic

Hallucination Frequency

Robustness

ZYC-ACCT-001Non-Agentic

Human-in-the-Loop

Accountability & Oversight

Risk scoring

Severity

Critical

score 4

Likelihood

Low

score 1

Risk

High

Control depth

4 mapped (below avg 5.2)

Framework refs

6/8 taxonomies

Framework mappings

Regulatory5/5

EU AI Act: Art. 9 (Risk management system — must identify and mitigate risks arising from emergent or unintended capabilities); Art. 15(3) (Cybersecurity — resilience against adversarial exploitation of model vulnerabilities); Art. 55(1)(a) (GPAI models with systemic risk — mandatory adversarial testing and red-teaming to surface emergent capabilities before deployment); Art. 72 (Post-market monitoring — continuous tracking of newly identified risks in deployed systems); Art. 73 (Reporting of serious incidents caused by emergent capability exploitation)
DORA: Art. 8 (Identification — mapping of ICT assets and novel threat vectors including AI-layer vulnerabilities); Art. 9(2) (Protection and prevention — controls against exploitation of undiscovered attack surfaces); Art. 10 (Detection — anomaly monitoring to identify exploitation of emergent capabilities in production); Art. 17 (ICT-related incident management process — response when emergent capability is actively exploited)
RBI: People First
NIST AI RMF: GOVERN 1.1: Legal/regulatory requirements; MAP 5.1: Impact assessment
ISO/IEC 42001: Clause 6.1.2: AI risk assessment criteria; Clause 5.2: AI policy

Risk taxonomies4/4

MIT AI Risk Repository: Domain 6 – Socioeconomic
IBM AI Risk Atlas: Impact on Human Agency
NIST ARIA: —
MAESTRO: L7 – Ecosystem

Security2/4

OWASP: —
MITRE ATLAS: —
Cisco AI Risk: — Not covered by Cisco AI Defense
Palo Alto AI Risk: — Not covered by Palo Alto Prisma AIRS

Capability‑pattern applicability

Classification

Foundational (inherited)

Agentic pattern

Foundational (inherited)

Capability pattern

Document Extraction; RAG; Summarization & Chat

Pattern scope

Pattern-agnostic (machinery)

Agent loop stage

Output

Why

Agent-machinery risk — realised in any deployment pattern; filter primarily by Agentic Pattern. [proposed]

Ownership & threshold

Owner (RACI)

AI Governance Office [A]

Threshold

Programme KRI within board-approved tolerance

Inherent vs residual

Inherent

High

Residual

2.4

High

Mapped controls: 4
Assumed control effectiveness: 40%

Modeled, not measured — design-stage; controls not yet tested (reduction capped 40%). Residual applies a uniform assumed reduction to the inherent score; it is not evidence of tested control performance.

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-005High

AI-Generated Information Manipulation

Search the Atlas

AI-Enabled Authority Centralisation

Description

Primary mitigations

Detection signals

Mitigating controls

Related risks in Societal & Economic Risk