Back to the MIT repository
6. Socioeconomic and Environmental2 - Post-deployment

Ability to automate jobs

The ability to automate jobs by AI models and systems can lead to significant job displacement, economic disruption, and social inequality.

Source: MIT AI Risk Repositorymit1045

ENTITY

2 - AI

INTENT

1 - Intentional

TIMING

2 - Post-deployment

Risk ID

mit1045

Domain lineage

6. Socioeconomic and Environmental

262 mapped risks

6.2 > Increased inequality and decline in employment quality

Mitigation strategy

1. Investment in Comprehensive Workforce Transition Programs: Strategically fund and implement national-scale reskilling and upskilling programs to prepare workers for new roles in the AI-augmented economy. These initiatives must prioritize foundational AI literacy, cross-disciplinary skills, and the development of inherently human capabilities such as critical thinking, creative problem-solving, and emotional intelligence, ensuring accessibility and equitable distribution across affected demographics. 2. Structural Reform of Economic and Social Safety Nets: Establish robust social insurance mechanisms designed for a highly automated labor market. This includes supporting policies for portable health and retirement benefits, creating dedicated worker retraining accounts, and extending or supplementing unemployment compensation (e.g., AI Adjustment Assistance) for individuals displaced by automation to facilitate career transitions without undue economic hardship. 3. Mandate Job Redesign and Organizational Transparency: Institute regulatory requirements compelling organizations to proactively redesign job roles by reallocating routine, automatable tasks to AI systems while redirecting human effort toward higher-value tasks that require emotional intelligence, ethical judgment, and complex human interaction. Simultaneously, enforce mandatory transparency regarding the use of AI in employment-related decisions, such as hiring, performance evaluation, and disciplinary actions, to safeguard against bias and ensure human oversight.