서브메뉴
검색
Artificial Intelligence and the Operationalization of SESTA-FOSTA on Social Media Platforms
Artificial Intelligence and the Operationalization of SESTA-FOSTA on Social Media Platforms
상세정보
- 자료유형
- 학위논문 서양
- 최종처리일시
- 20260202105106
- ISBN
- 9798297601192
- DDC
- 301
- 저자명
- Barreto, Renata.
- 서명/저자
- Artificial Intelligence and the Operationalization of SESTA-FOSTA on Social Media Platforms
- 발행사항
- [Sl] : University of California, Berkeley, 2025
- 발행사항
- Ann Arbor : ProQuest Dissertations & Theses, 2025
- 형태사항
- 132 p
- 주기사항
- Source: Dissertations Abstracts International, Volume: 87-04, Section: B.
- 주기사항
- Advisor: Obasogie, Osagie;Lee, Taeku.
- 학위논문주기
- Thesis (Ph.D.)--University of California, Berkeley, 2025.
- 초록/해제
- 요약This dissertation investigates how AI content moderation systems reflect and intensify sociotechnical inequalities, particularly in the wake of U.S. regulatory shifts like SESTA-FOSTA. Through three interdisciplinary studies, it traces the entanglement of law, algorithmic enforcement, and marginalized users' online experiences. The first paper audits the open_nsfw image classifier, using an Internet Archive dataset and embeddingbased computer vision to reveal systematic gender- and sexuality-based bias. It shows how regulatory pressure led Tumblr to adopt a model that disproportionately flagged queer, femme, and artistic content, embedding mainstream moral norms into machine learning systems. The second paper analyzes a large survey of Instagram creators- primarily BIPOC, LGBTQ+, and disabled-collected with Salty. It documents disparities in perceived takedowns, shadowbanning, and suspensions, especially post-SESTAFOSTA, and shows how users interpret and resist opaque algorithmic governance. The third paper provides a sociolegal analysis of SESTA-FOSTA's impact on CDA 230, mapping case law and demonstrating how courts have struggled with its ambiguity and symbolic scope. Using theories of legal endogeneity and delegated enforcement, it critiques how the law outsources regulatory power to platforms, often harming vulnerable users. Together, these studies expose the mechanics and consequences of embedding law into code, and they call for external audits, community-informed regulation, and governance frameworks that address harm without further marginalization.
- 일반주제명
- Sociology
- 일반주제명
- Computer science
- 일반주제명
- Law
- 일반주제명
- Web studies
- 키워드
- AI ethics
- 키워드
- AI safety
- 키워드
- Algorithmic bias
- 기타저자
- University of California, Berkeley Jurisprudence & Social Policy
- 기본자료저록
- Dissertations Abstracts International. 87-04B.
- 전자적 위치 및 접속
- 로그인 후 원문을 볼 수 있습니다.
MARC
008260126s2025 us c eng d■001000017359348
■00520260202105106
■006m o d
■007cr#unu||||||||
■020 ▼a9798297601192
■035 ▼a(MiAaPQ)AAI32236627
■040 ▼aMiAaPQ▼cMiAaPQ
■0820 ▼a301
■1001 ▼aBarreto, Renata.
■24510▼aArtificial Intelligence and the Operationalization of SESTA-FOSTA on Social Media Platforms
■260 ▼a[Sl]▼bUniversity of California, Berkeley▼c2025
■260 1▼aAnn Arbor▼bProQuest Dissertations & Theses▼c2025
■300 ▼a132 p
■500 ▼aSource: Dissertations Abstracts International, Volume: 87-04, Section: B.
■500 ▼aAdvisor: Obasogie, Osagie;Lee, Taeku.
■5021 ▼aThesis (Ph.D.)--University of California, Berkeley, 2025.
■520 ▼aThis dissertation investigates how AI content moderation systems reflect and intensify sociotechnical inequalities, particularly in the wake of U.S. regulatory shifts like SESTA-FOSTA. Through three interdisciplinary studies, it traces the entanglement of law, algorithmic enforcement, and marginalized users' online experiences. The first paper audits the open_nsfw image classifier, using an Internet Archive dataset and embeddingbased computer vision to reveal systematic gender- and sexuality-based bias. It shows how regulatory pressure led Tumblr to adopt a model that disproportionately flagged queer, femme, and artistic content, embedding mainstream moral norms into machine learning systems. The second paper analyzes a large survey of Instagram creators- primarily BIPOC, LGBTQ+, and disabled-collected with Salty. It documents disparities in perceived takedowns, shadowbanning, and suspensions, especially post-SESTAFOSTA, and shows how users interpret and resist opaque algorithmic governance. The third paper provides a sociolegal analysis of SESTA-FOSTA's impact on CDA 230, mapping case law and demonstrating how courts have struggled with its ambiguity and symbolic scope. Using theories of legal endogeneity and delegated enforcement, it critiques how the law outsources regulatory power to platforms, often harming vulnerable users. Together, these studies expose the mechanics and consequences of embedding law into code, and they call for external audits, community-informed regulation, and governance frameworks that address harm without further marginalization.
■590 ▼aSchool code: 0028.
■650 4▼aSociology
■650 4▼aComputer science
■650 4▼aLaw
■650 4▼aWeb studies
■653 ▼aAI ethics
■653 ▼aAI safety
■653 ▼aAlgorithmic bias
■653 ▼aContent moderation
■653 ▼aPlatform governance
■653 ▼aSociotechnical systems
■690 ▼a0626
■690 ▼a0984
■690 ▼a0398
■690 ▼a0800
■690 ▼a0646
■71020▼aUniversity of California, Berkeley▼bJurisprudence & Social Policy.
■7730 ▼tDissertations Abstracts International▼g87-04B.
■790 ▼a0028
■791 ▼aPh.D.
■792 ▼a2025
■793 ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17359348▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.
