RightsCon Toronto 2018

Session Emcee: Isaac Stone Fish

Less Hate in Politics! Machine Learning and Interventions as Tools to Mitigate Online Hate Speech in Political Campaigns (Open Knowledge Finland)

Speakers: Teemu Ropponen

The session is a project presentation and a call for action on similar iniatives - though we would love to be part of another workshop or demo session, if there were matching to be done. The session presents questions, tools, practical experience up to government level politics.

Discriminating, hateful speech online has become a pressing problem in the societies. What is challenging is that ‘hate speech’ has come to mean a variety of speech acts and other ill-behaviours online, ranging from penal criminal acts to speech which is uncivil and disturbing, but yet to be tolerated. This definitional difficulty is further abused in claims that any limitations of hate speech endanger people’s right to freedom of expression.

Hate speech has been criminalized in many countries. However, the automatic identification of hate speech is limited by lacking tools. While there are few open tools to detect hate speech (e.g. Conversation AI), they are only capable of handling English language. However, these language challenges can be solved with various supervised text mining techniques, as long as the laborious face of developing a gold standard for detection is done properly.

We present a project where the social media activity of candidates was monitored during the Finnish municipal election campaigns in April 2017. The monitoring project was initiated by Open Knowledge Finland, together with the Non-Discrimination Ombudsman (a governmental body to prevent and tackle discrimination), The Finnish League for Human Rights, software company Futurice, and three academic research projects. The pilot project included technical infrastructures to automatically filter potential hate speech from large social media monitoring data - over 7000 public Facebook and Twitter accounts of candidates, available as open data.

Later, populist True Finns party politicians reacted, and felt that his was a Big Brother watching the candidates. This went as far the Finnish Parliament, for the Minister of Justice to answer to the populist True Finns’s question about the project.

Love is a battlefield. Experiences studying life hacks to counter online violence against women and LGBT activists in Latin America (NVA Labs)

Speakers: Lu Ortiz

We will share the findings of the study we conducted to collect life hacks and collective knowledge from feminist collectives and LBGT groups to counter and combat online violences in Latin America. The testimonies reflect strength, innovation, creative forms of risk assessment and poise.

Digital Rights are LGBTQ Rights: or What to do When your Content Turns Against You (GLAAD)

Speakers: Jim Halloran

Last year, the media advocacy organization GLAAD uncovered an online effort to weaponize LGBTQ content against itself. These homophobic and transphobic attacks gamified social media algorithms with such scale and repetition that they began to influence the AI behind some the largest online platforms.

Since then, GLAAD’s Digital program has been leading a fight against both bad code, like toxic AI training data, and bad actors, such as a Stanford University professor who dangerously claims to have built AI that can detect sexual orientation.

Learn the challenges and triumphs in protecting the digital lifeline for LGBTQ people who are increasingly being left behind by AI – and need to have a seat at the table in building inclusive AI.

A ToT model for Safe Sisters (Internews)

Speakers: Natasha Msonza & Haley Slafer

This is a presentation of findings, reflections and feedback collected from an experiment undertaken as part of a Training of Trainers learning methodology. It involved 12 at-risk women human rights activists from Rwanda, Uganda, Kenya, Tanzania and Burundi and used a combination of self-study, tool practice, mentorship, workshops, peer sharing, and small grants. This cohort of technologists joined forces to take on today’s digital challenges for women human rights defenders.