A leaked internal document reveals a major tech company’s plans to overhaul its content moderation policies, potentially allowing more user-generated material to bypass automated filters. The proposed changes, described as a shift toward "contextual understanding" rather than strict keyword blocking, could reshape how harmful or misleading content is handled online. Critics warn the move risks amplifying misinformation and extremist material, while supporters argue it may improve free expression and reduce over-censorship. The document suggests the company is testing new AI tools to refine its approach before a broader rollout.