MIT AI Risk Repository · Risk Sub-Category · 24.10.01
Entrenchment and exacerbation of existing inequalities
Category: Access and Opportunity risks
Description
"The most serious access-related risks posed by advanced AI assistants concern the entrenchment and exacerbation of existing inequalities (World Inequality Database) or the creation of novel, previously unknown, inequities. While advanced AI assistants are novel technology in certain respects, there are reasons to believe that – without direct design interventions – they will continue to be affected by inequities evidenced in present-day AI systems (Bommasani et al., 2022a). Many of the access-related risks we foresee mirror those described in the case studies and types of differential access.
From The Ethics of Advanced AI Assistants (Gabriel2024), as extracted by the MIT AI Risk Repository (CC BY 4.0).
Classification
- Causal entity
- Human
- Intent
- Unintentional
- Timing
- Post-deployment
Subdomain definition: AI-driven concentration of power and resources within certain entities or groups, especially those with access to or ownership of powerful AI systems, leading to inequitable distribution of benefits and increased societal inequality.
Real-world incidents in this subdomain
- LLM Scrapers Allegedly Target Multiple Open Source Projects Disrupting the FOSS Ecosystem
- Coupang Allegedly Tweaked Search Algorithms to Boost Own Products
- Amazon Allegedly Tweaked Search Algorithm to Boost Its Own Products
- Gmail’s Inbox Sorting System Reportedly Reduced Visibility of Political Emails and Campaign Calls-to-Action
- Google Fined for Changing Shopping Algorithms in EU to Favor Own Service
- Amazon India Allegedly Rigged Search Results to Promote Own Products
How other frameworks describe this risk
Other entries from Gabriel2024
- Capability failures
- Lack of capability for task
- Difficult to develop metrics for evaluating benefits or harms caused by AI assistants
- Safe exploration problem with widely deployed AI assistants
- Goal-related failures
- Misaligned consequentialist reasoning
- Specification gaming
- Goal misgeneralisation
- Deceptive alignment
- Malicious Uses
- Offensive Cyber Operations (General)
- AI-Powered Spear-Phishing at Scale