Following #MotherLanguageDay, we’re reminded that the "AI gap" harms linguistic diversity. To fix it, we must listen.
Today, with Hiperderecho, we’re releasing 2 new videos on content moderation in Quechua.🧵👇🏾
Quechua: www.youtube.com/watc...
Spanish:
Supported by Internet Society Foundation, these videos are a tool for change.
Watch in Quechua: www.youtube.com/watc...
Watch in Spanish: www.youtube.com/watc...
Read the full report: cdt.org/insights/mod...
It’s not every day that UX researcher whistleblowers testify before the Senate. Yesterday, two former Meta researchers, Jason Sattizahn and Cayce Savage, shared their concerns with safety research for Meta VR products, and more broadly within Meta. So how does UX research move on from here? 🧵
To make language technologies better, AI systems need to be built with native speakers and language experts, some of whom we highlight in this work, others include folks at Whose Knowledge, Mozilla, Karya, Masakhane, Te Hiku and more:
cdt.org/insights/con....
Our own study of multilingual AI finds that these systems are trained on mostly English-lang data, machine translated speech, or data scraped from the web which doesn't represent how people speak. Translation is an art form. AI is creating six-fingered mimicry of language: cdt.org/insights/los...
🗣️ Interpretation in EN/SW/AR/ES
🕘 9:00–10:30 am ET
cdt.org/event/interr...
Most content moderation systems weren’t built with the Global South in mind. CDT’s new research shows how they fail in Tamil, Kiswahili, Quechua & more.
Join our Aug 7 webinar for findings, recommendations, and global voices on language equity in tech.
Most content moderation systems weren’t built with the Global South in mind. CDT’s new research shows how they fail in Tamil, Kiswahili, Quechua & more.
Join our Aug 7 webinar for findings, recommendations, and global voices on language equity in tech.
President Trump's attempt to fire the Democratic members of the #FederalTradeCommission today is outrageous. The FTC is now the latest agency to see its Dem commissioners fired without cause, joining the Privacy & Civil Liberties Oversight Board, EEOC, Federal Election Commission, NLRB & more.
This is going to be SUCH a great discussion!
Join me alongside Centre for Internet & Society, Paradigm Initiative, and Hiperderecho at #RightsCon2025 next week talking equitable content moderation! More sessions on enhancing multilingual AI and human rights implications of LLMs here: cdt.org/insights/cdt...
Ripoti hii inaangazia jinsi mbinu za sasa za usimamizi wa maudhui zinavyoshindwa kukidhi mahitaji ya wazungumzaji wa Kiswahili zaidi ya milioni 100.
Soma ripoti kamili hapa: cdt.org/insights/mod...
Je, #UsimamiziWaMaudhui ya kiswahili kwenye mitandao ya kijamii unawaathiri vipi wazungumzaji wa Kiswahili? Tumeshirikiana na Paradigm Initiative kuandaa ripoti hii iliyotafsiriwa kwa lugha ya kiswahili. cdt.org/insights/mod...
Great report from my colleagues @aliyabhatia.bsky.social and @mluria.bsky.social. So timely!
This #BlackHistoryMonth, we’re reflecting on the work needed to center racial justice in technology policy. Regardless of political winds, it’s critical to address systemic biases & inequities that can disproportionately harm Black communities.
How is #ContentModeration impacting Kiswahili speakers? We partnered with Paradigm Initiative for this translated report. Now available in Kiswahili, “Moderating Kiswahili Content on Social Media” dives into how current content moderation practices fail to meet the needs of 100M+ Kiswahili speakers.
CDT is joined by the AAPD, AAAJ, @ourcommoncause.bsky.social, @glaad.bsky.social, @hrc.org, @lawyerscomm.bsky.social, Muslim Public Affairs Council, NAACP LDF, National Black Justice Coalition, @natfairhouse.bsky.social , National Urban League
Today, @cendemtech.bsky.social + partner orgs sent an open letter to Mark Zuckerberg expressing grave concerns about changes to Meta’s content moderation policies which threaten marginalized groups and undermine free expression. Here’s what you need to know: cdt.org/insights/let...
The report also delves into privacy and human rights implications of these tools. Transparency, bias, and unintended impacts on users must be carefully managed. Read more on how platforms can enhance trust & safety responsibly: cdt.org/insights/rea... #SafeOnline #ChildOnlineSafety
Last week, CDT released a report that explores the impact of AI-powered hiring tools on disabled job applicants. Learn more about the risks, challenges, and recommendations for building fair and inclusive hiring systems. cdt.org/insights/scr...
٣/ في هذا التقرير٬ نلقي الضوء علي التحديات المتعلقة بالإشراف علي المحتوي المتعلق بلغات الدول التي تقع في الجنوب. قد درسنا ايضا التحيزات التي اثرت علي مستخدمي وسائل التواصل الإجتماعي.
We dive into the challenges of moderating Global South languages & explore biases that impact users.
٢/لقد تعاونا مع DigitalCitizenship وتلقينا دعم من ISOC Foundation حتي ننشر هذا التقرير
We worked in partnership with DigitalCitizenship and with support from ISOC Foundation, to release this report. cdt.org/insights/mod...
🚨NEW REPORT:🚨 We’re excited to announce the release of the Arabic language version of our report, “Moderating Maghrebi Arabic Content on Social Media.” cdt.org/insights/mod... #ContentModeration
🚨 تقرير جديد 🚨
١/يسعدنا ان نعلن عن نشر الترجمة العربية للتقرير " إدارة المحتوى
العربي المغاربي على وسائل التواصل الاجتماعي" #إشراف_المحتوي cdt.org/insights/mod...
The lack of representation of Maghrebi Arabic speakers in developing automated moderation systems significantly reduces the accuracy and fairness of moderation decisions. Check out the report for more insights: cdt.org/insights/mod...
The report reveals that content moderators reviewing Arabic content are often assigned content from any country in the region regardless of their native dialect, leading to errors and inconsistencies in moderation. cdt.org/insights/mod...
Our research, led by Mona Elswah, found that Maghrebi Arabic users use “algospeak” as a creative tactic to bypass moderation algorithms & users often mass report content to compensate for ineffective reporting mechanisms. cdt.org/insights/mod...
Our report focusing on Maghrebi Arabic dialects is the first in an ongoing series investigating content moderation biases in the Global South. cdt.org/insights/mod...
NEW RESEARCH: CDT’s report (in partnership with DigitalCitizenship & funded by @internetsociety.bsky.social), “Moderating Maghrebi Arabic Content on Social Media,” analyzes the challenges of moderating Global South languages. cdt.org/insights/mod...