UN Report Cites Starling Lab Submission on AI Surveillance Risks
The UN Special Rapporteur on Counter-Terrorism published their position paper, “Protecting Human Rights while Using Artificial Intelligence to Counter Terrorism,” warning that AI-driven mass surveillance can lead to the “suppression of human rights defenders, activists, dissidents, minorities and journalists.”
Notably, the report cites Starling Lab’s submission when addressing these “freedom of expression concerns.”
In our contribution, we warned of a “pendulum effect” where the very tools designed to establish truth could be weaponized. We argue that “technology intended to protect human rights by verifying truth can, if implemented without ethical guardrails like opt-in principles and privacy-preserving designs, be used to violate them.” From the submission:
“(…) It is critical to address a potential ‘pendulum effect’ that is a direct consequence of the threats posed by AI. In response to these AI-driven threats, counter-terrorism demands may lead to policies that mandate authenticity infrastructure. This creates a significant risk that tools designed to establish authenticity could be repurposed for unwarranted surveillance and invasion of privacy. Therefore, any conversation about safeguards must acknowledge this dilemma: technology intended to protect human rights by verifying truth can, if implemented without ethical guardrails like opt-in principles and privacy-preserving designs, be used to violate them.”
The UN’s acknowledgement reinforces our core message: To prevent the misuse of AI, we must “proactively design our systems for authenticity.” Protecting rights in this new era requires more than policy – it requires technical accountability and cryptographic proofs of provenance that support, rather than undermine, civil liberties. From the submission:
“In summary, we urge a paradigm shift in how we approach digital information in the age of AI. Instead of a reactive posture focused on detecting forgeries, we must proactively design our systems for authenticity and long-term preservation.”
“Our central principle: Authenticate Upstream, Not Downstream… This secure foundation allows for the creation of stronger, more contextualized, and self-verifiable assets. By embedding provenance metadata at the point of capture, we create a secure digital paper trail for media – directly protecting several fundamental human rights, especially in the high-stakes context of countering terrorism.”