Authoritarian Surveillance, Censorship, and Use (General)
While new technologies like advanced AI assistants can aid in the production and dissemination of decision-guiding information, they can also enable and exacerbate threats to production and dissemination of reliable information and, without the proper mitigations, can be powerful targeting tools for oppression and control. Increasingly capable general-purpose AI assistants combined with our digital dependence in all walks of life increase the risk of authoritarian surveillance and censorship. In parallel, new sensors have flooded the modern world. The internet of things, phones, cars, homes, and social media platforms collect troves of data, which can then be integrated by advanced AI assistants with external tool-use and multimodal capabilities to assist malicious actors in identifying, targeting, manipulating, or coercing citizens.
ENTITY
1 - Human
INTENT
1 - Intentional
TIMING
2 - Post-deployment
Risk ID
mit388
Domain lineage
4. Malicious Actors & Misuse
4.1 > Disinformation, surveillance, and influence at scale
Mitigation strategy
1. **Implement Privacy-Preserving Machine Learning (PPML) Architectures:** Mandate the use of techniques such as Differential Privacy and Secure Federated Learning during the AI's development to ensure that individual user data cannot be aggregated or de-anonymized at scale, thereby structurally inhibiting the AI's utility as a mass surveillance or targeting tool. 2. **Establish and Enforce Transparent, Auditable Use Policies:** Draft and publicly commit to stringent ethical guidelines that explicitly prohibit the deployment of the AI assistant for state-sponsored surveillance, content censorship, or political coercion, coupled with robust, independent auditing mechanisms to monitor API access, data input, and deployment contexts for non-compliance. 3. **Conduct Rigorous Adversarial Misuse and Data Integration Red-Teaming:** Execute continuous, specialized red-teaming simulations that focus on the AI's ability to integrate data from external, multimodal sources (e.g., IoT, social media feeds) to identify, track, or manipulate populations, ensuring vulnerabilities are identified and patched before systems are deployed in sensitive domains.