VIDEO
HIGHLIGHT
Log InSign up
Individual skills to address harmful practices and concluding initiatives
SummaryTranscriptChat

Individual skills to address harmful practices and concluding initiatives

Welcome and Introduction

The introduction to the course on freedom of expression, artificial intelligence, and elections organized by the Knight Center for Journalism in the Americas, UNESCO, UNDP, and the Electoral Assistance Division of the United Nations.

Course Introduction

  • The course is a collaboration between the Knight Center for Journalism in the Americas, UNESCO, UNDP, and the Electoral Assistance Division of the United Nations.
  • Albertina Peta is an electoral expert at UNESCO and leads this course on freedom of expression and journalist safety.

Verifying AI-Generated Content

Techniques for verifying and detecting AI-generated content focusing on manual detection methods.

Verifying AI Content

  • Manual detection involves analyzing signs of AI-generated content such as imitating human writing styles and generating realistic audios, photos, and videos.
  • Glitches in AI models can aid manual detection as they may still allow identification of synthetic material.
  • Analyzing metadata can help detect AI-generated content by checking for markers left by authors or discrepancies in date/time stamps.
  • Manual analysis includes looking for glitches like repetitive text or unusual body parts in photos/videos that indicate AI generation.

Detecting Synthetic Content with AI Tools

Discusses using AI tools for automatic detection to support identifying synthetic content.

Automatic Detection Tools

  • Various tools like GPT Z zero GPT originality are used for detecting AI-generated text while others like OpticAI analyze images.
  • Deepwear or FactCatcher are employed for video analysis while Microsoft's authenticator tool examines photos and videos.
  • Emphasizes using multiple detection methods due to evolving nature of AI detection tools to ensure accurate verification of information.

Media and Information Literacy Interventions

This section discusses the importance of media and information literacy interventions, distinguishing between long-term and short-term strategies.

Media and Information Literacy Interventions

  • Long-term interventions involve integrating MIL into educational curricula, community engagement programs, and advocating for supportive policies to build sustainable public resilience to information pollution.
  • Short-term interventions focus on adult education, empowering electoral stakeholders through workshops, and conducting public awareness campaigns to enhance media literacy skills among voters and citizens.

Challenges in Media and Information Literacy during Elections

This part highlights the challenges faced in media and information literacy during electoral processes.

Main Challenges

  • The rapid evolution of sophisticated artificial intelligence methods poses a challenge in keeping knowledge updated with technology advancements.
  • Human-based challenges arise from effectively applying the toolkit provided by media and information literacy to make informed decisions during elections.

Fostering Civic Agency Through Media Literacy

This segment explores how media literacy initiatives aim to foster civic agency among individuals.

Civic Agency Development

  • Civic agency refers to individuals' capacity to participate in public life, exercise rights, contribute to communities, engage in democratic processes, advocate for social change, promote peace, tolerance, and curb polarization.

Advancements in Media Literacy Responses

Discusses advancements in media literacy responses across various policy areas.

Advancements

  • Information search and access are crucial advancements where policies ensure quality data control while responses empower users for diverse search results with gender perspectives.

Guidelines for Governance of Digital Platforms

This section discusses the UNESCO guidelines for the governance of digital platforms, emphasizing inclusivity and empowerment.

Multistakeholder Process

  • The guidelines were developed through a multistakeholder process involving representatives from over 130 countries.
  • More than 10,000 comments were received globally, ensuring inclusivity across diverse backgrounds and marginalized groups.
  • Emphasis on empowering users, promoting cultural diversity, social inclusion, global citizenship, gender equality, and women's empowerment.

Government Responsibilities

  • Governments should prioritize media literacy and online safety for vulnerable groups.
  • Collaboration with experts is essential to educate users about their rights.
  • Digital platforms must empower users and enhance digital literacy with a focus on vulnerable groups.

Role of Journalists

  • Journalists play a crucial role in contributing to transparency in elections by fact-checking information.
  • Building trust in media is vital through accurate reporting and verification processes.

Ethical Impact Assessments for AI Systems

This part delves into ethical impact assessments for AI systems, highlighting the importance of identifying positive and negative impacts throughout the system's life cycle.

Importance of Ethical Impact Assessments

  • Ethical impact assessments help identify both positive and negative impacts of AI systems on human rights and fundamental freedoms.
  • These assessments are crucial to address biases within AI systems that may arise from data sources.

Reasons for Conducting Assessments

  • Identifying weaknesses such as biases is a functional requirement to improve AI systems.
  • Organizational imperatives include enhancing governance systems and ensuring regulatory compliance when using AI technologies.

Benefits of Assessments

  • Individuals benefit from better protection of their rights through ethical impact assessments on AI systems.
  • Promotes social good by fostering innovation, sustainable development, and enhancing trust in technologies.

Tools for Ethical Impact Assessment

Introduces tools like the UNESCO ethical impact assessment tool focusing on scoping questions and alignment with UNESCO principles.

Scoping Questions

  • Scoping questions help understand an AI system's objectives, context of deployment, project governance setup involvement by stakeholder groups.

Alignment with Principles

  • The tool aligns the AI system with UNESCO principles including safety, security fairness privacy transparency among others.

Ethical Impact Assessment of AI Systems

In this section, the importance of ethical impact assessments for AI systems is discussed, emphasizing accountability, transparency, and alignment with principles such as non-discrimination.

Importance of Ethical Impact Assessments

  • Ethical impact assessments help in fixing accountability and ensuring transparency in deploying AI systems.
  • Alignment with principles like non-discrimination involves testing algorithms with diverse stakeholder groups and analyzing training and testing data for biases.
  • Conducting an ethical impact assessment prompts reflection on the impacts of AI systems, advocating for a multi-stakeholder approach involving potential users in the assessment process.

Coalitions for Addressing Risks in Information Ecosystem

This part focuses on the role of coalitions in mitigating risks associated with AI in the information ecosystem, emphasizing coordination, leveraging diverse skill sets, and regulatory efforts.

Role of Coalitions

  • Coalitions play a crucial role in addressing risks by coordinating responses efficiently and leveraging diverse skill sets to bridge power imbalances between national institutions and global companies.
  • They support governance by balancing increased access to AI while managing associated risks, impacting regulatory efforts through evidence-based decision-making and monitoring legislative processes.

Activities to Mitigate Risks in Information Ecosystem

This segment delves into specific activities such as media literacy programs, content moderation tools, social listening strategies, and strategic communication to mitigate risks like disinformation and violence.

Mitigation Activities

  • Media literacy programs coupled with civic education enhance citizen resilience against online disinformation and empower them to discern trustworthy information sources.
  • Tools for detecting AI-generated content aid in combating polarizing narratives on social media platforms that can incite violence against vulnerable groups.

Enhancing Transparency Through Strategic Communication

The focus here is on strategic communication strategies that electoral authorities can employ to boost trust in electoral processes through accurate information dissemination and stakeholder engagement.

Strategic Communication Measures

Activity Commitments and Technologies

The speaker discusses the importance of voluntary commitments by stakeholders like political parties, candidates, officials, and media in using technologies responsibly to enhance electoral processes. Monitoring and enforcing these commitments can be challenging but are crucial for a healthy ecosystem.

Voluntary Commitments for Electoral Processes

  • Stakeholders such as political parties, candidates, officials, and media should make voluntary commitments to use technologies responsibly.
  • Building coalitions to monitor commitment implementation and consulting stakeholders extensively can improve adherence.
  • Training media, election personnel, and AI practitioners is essential for better understanding and implementation of commitments.

Ethical Considerations in AI Projects

The speaker emphasizes the need for ethical testing in all AI projects involving electoral processes. Adhering to digital principles outlined by organizations like UNDP is crucial for ensuring integrity in information control institutions.

Ethical Testing and Digital Principles

  • Conducting ethical tests for all AI projects related to elections is recommended.
  • Adherence to digital standards like those from UNDP is vital as CSOs increasingly utilize AI systems.

Innovative Approaches with AI Systems

The discussion focuses on combining existing governance support activities with innovative approaches using AI systems. Emphasizing the importance of continuous assessment, support measures enhancement, and innovation in utilizing AI reasoning capacities.

Integration of Governance Support with Innovative AI Systems

  • Combining traditional governance support activities with innovative approaches using AI systems is critical.
  • Continuous efforts are needed to assess, enhance support measures, innovate due to emerging capabilities of AI systems.

Acknowledgments and Gratitude

Expressing gratitude towards collaborators who contributed significantly to the course's content development. Acknowledging the dedication of colleagues from various organizations who played key roles in shaping the course material.

Recognition and Appreciation

  • Acknowledgment of collaborative efforts involving over 20 individuals working tirelessly under exceptional coordination.