Meta considering replacement of humans with AI to assess privacy and societal risks

Meta considering replacement of humans with AI to assess privacy and societal risks
—————————————–
Meta is shifting from human-led to AI-driven risk assessments for its platforms, automating up to 90% of evaluations related to privacy, safety, and content integrity, Npr.org reported. This change means that critical updates to Instagram, WhatsApp, and Facebook will largely be approved by AI systems rather than human reviewers.

Previously, human evaluators assessed potential risks such as privacy violations, harm to minors, and the spread of misinformation before product launches. The new process empowers product teams to make their own risk judgments, with AI providing instant decisions based on questionnaires about the projects. Human reviews will only occur for novel or complex cases, and teams can request manual assessments if needed.

Meta states the automation aims to streamline decision-making and speed up product releases while maintaining human oversight for complex issues. However, concerns have been raised internally and externally that this approach could increase risks by reducing rigorous scrutiny and relying on engineers who are not privacy experts. Critics warn that faster launches may lead to overlooked problems and negative consequences.

The company continues to audit AI decisions and maintains stricter oversight for products in the European Union, complying with regulations like the Digital Services Act. This shift follows Meta’s broader strategy to leverage AI for faster innovation amid competition and reflects a loosening of previous content moderation policies and fact-checking programs.