Background:

At the OSCE Secretariat, the Transnational Threat Department’s Action against Terrorism Unit (TNTD/ATU) is the focal point and a resource hub for the OSCE’s 57 participating states, field operations and independent institutions, and helps support and implement the OSCE’s counter terrorism commitments. OSCE participating States have identified countering the use of the Internet for terrorist purposes (CUITP) as one of the strategic focus areas of OSCE counter-terrorism efforts. The OSCE Consolidated Framework for the Fight against Terrorism (PC.DEC/1063, 2012) provides the overarching framework for OSCE support in this area. The OSCE Ministerial Declaration on Preventing and Countering Violent Extremism and Radicalization that Lead to Terrorism (MC.DOC/4/15) underscores the importance of comprehensive, multi-stakeholder and human rights-compliant approaches, including public-private partnerships, the use of social media to counter violent extremist messaging, youth engagement, gender-sensitive approaches, tolerance and non-discrimination.

The OSCE’s work in this field is also informed by participating States’ commitments to intensify efforts to address the use of the Internet for terrorist purposes, to identify possible strategies to combat this threat while ensuring respect for relevant international human rights obligations and standards, and to promote capacity-building and the exchange of information and good practices at national, regional and OSCE-wide levels.

Within this mandate, TNTD/ATU implements the ExB Project ‘CLEAR: Critical Thinking, Information and Media Literacy, Ethical artificial intelligence (AI), and Resilience for Preventing and Countering Violent Extremism and Radicalization that Lead to Terrorism’. The project aims to strengthen comprehensive and human rights-compliant approaches to preventing and countering violent extremism and radicalization that lead to terrorism (P/CVERLT) that manifests in, and stems from, online environments across the OSCE area.

Project CLEAR combines three mutually reinforcing components: (1) building media and information, AI and digital literacy capacities to enhance community resilience against violent extremist and terrorist online content; (2) strengthening government and non-governmental organisations’ awareness and preparedness regarding the responsible and human rights-compliant use of AI and evolving technologies in both P/CVERLT and counter-terrorism; and (3) improving cross-border co-operation and institutional capacities for requesting, handling and exchanging electronic evidence essential for countering terrorist misuse of digital platforms.

As part of the AI-related line of work under Project CLEAR, TNTD/ATU intends to support participating States and other stakeholders in understanding AI-enabled threats, including AI-generated content, deepfakes, automated amplification, algorithmically targeted influence operations and the misuse of evolving technologies by violent extremist and terrorist actors. At the same time, the project promotes the responsible, ethical and human rights-compliant use of AI-supported tools for prevention and counter-terrorism purposes, with due attention to legality, necessity, proportionality, meaningful human oversight, accountability, transparency, data protection, auditability, security safeguards and the do-no-harm principle.

Project activities may include AI literacy workshops, practical training courses, training-of-trainers, policy dialogues, tabletop exercises and tailored capacity-building activities for different categories of participants, including law enforcement and justice and security actors, policymakers, national training institutions, civil society, youth, educators, media representatives, researchers and, where relevant, technology-sector stakeholders. Training activities will be tailored to the audience, ranging from introductory and awareness-raising sessions to more technical modules on AI, machine learning (ML), deep learning (DL), large language models (LLMs), responsible AI governance and practical use-case design.

Objective of Assignment:

The objective of this roster call is to invite qualified candidates with demonstrated expertise in artificial intelligence (AI), machine learning (ML), deep learning (DL), generative AI and responsible AI governance to join the OSCE roster of experts supporting Project CLEAR and other related activities in this field organized by the OSCE.

Once admitted to the roster, experts may be considered for specific consultancies to support the design, adaptation and delivery of training courses, workshops, tabletop exercises and policy or technical guidance on the responsible and human rights-compliant use of AI in the context of P/CVERLT and counter-terrorism. Shortlisted roster members will be contacted regarding specific assignments and requested to confirm their interest and availability. Admission to the roster may involve interviews and, where appropriate, a written or practical test.

Duration of Assignment:

This is a roster call. Duration of the assignment will depend on the tasks, deliverables, location and modality of each specific consultancy assigned under Project CLEAR or related TNTD/ATU activities.

Tasks and Responsibilities:

The OSCE is looking for professionals with demonstrated and in-depth expertise in AI/ML/DL and the ability to translate complex technical issues into practical, accessible and operationally relevant training content for different categories of stakeholders. Depending on the specific assignment, roster experts may be requested to:

• Develop, update or tailor training materials, slides, practical examples, demonstrations, case studies and exercises on AI/ML/DL, generative AI, LLMs, model families, prompting and safe use of AI tools, tokens, embeddings, inference, quantization, hallucinations, local and open-weight models, closed environments, vector databases and retrieval-augmented generation (RAG).
• Explain AI/ML/DL concepts in clear, simple and audience-appropriate language, including for non-technical or mixed groups such as law enforcement practitioners, policymakers, civil society representatives, educators, media professionals and youth participants.
• Support training participants in translating manual workflows and operational needs into AI/ML-relevant problem statements and in understanding the technical, organizational, ethical and human rights implications of possible AI-supported use cases.
• Design and facilitate practical demonstrations, group exercises and tabletop exercise inputs, including the use of synthetic or non-sensitive datasets, demo workflows and scenario-based discussions, while ensuring that no sensitive operational data is used unless specifically authorized under OSCE rules and applicable safeguards.
• Cover, as relevant, AI-assisted drafting and summarization, multilingual information processing, image and text analysis, local models and computing requirements, internal knowledge assistants, RAG, prompt injection and other AI security risks, model evaluation, operational validation and institutional readiness.
• Integrate responsible-use considerations throughout all materials and sessions, including human oversight, documentation, auditability, accountability, data protection, security safeguards, non-discrimination, gender sensitivity, do-no-harm and alignment with international human rights standards.
• Advise the OSCE project team on the technical feasibility of planned hands-on elements, including infrastructure needs, software requirements, connectivity constraints, computing requirements, data protection considerations and safe demonstration options.
• Co-operate with other experts, OSCE field operations, institutions and project partners as required, including by adapting content to regional or national contexts in Central Asia, South-Eastern Europe or other OSCE areas.
• Prepare brief written inputs, training summaries, lessons learned, technical notes or recommendations following activities, including observations on participants’ main questions, capacity needs and possible follow-up work.

Necessary Qualifications:

• Advanced university degree in computer science, data science, artificial intelligence, machine learning, applied mathematics, engineering or another relevant field; academic or teaching experience would be an asset;
• A minimum of 7 years of relevant professional experience, preferably in the field of machine learning, deep learning, data science or related technical fields, including practical experience with AI/ML model development, deployment, testing or evaluation;
• Strong knowledge of machine learning and deep learning concepts, including supervised and unsupervised learning, classification, clustering, model training and inference, model evaluation, natural language processing, embeddings and generative AI;
• Practical experience with large language models, open-weight models, local deployment options, RAG, vector databases, prompt engineering, synthetic or demo data and AI security risks would be a strong asset;
• Knowledge of ethical and responsible AI, AI governance, human oversight, auditability, data protection, security safeguards and the rule of law, preferably in law enforcement, public-sector, justice, security, counter-terrorism, P/CVERLT or other sensitive operational contexts;
• Experience in designing and delivering trainings, workshops or capacity-building activities on AI/ML/DL for non-technical or mixed operational audiences, including the ability to explain complex technical issues in clear and accessible language;
• Experience with Python and common AI/ML libraries or tools such as scikit-learn, TensorFlow, PyTorch, Hugging Face, Ollama or similar tools would be an asset;
• Previous experience providing international capacity-building assistance, preferably with the OSCE, the United Nations or other international or regional organizations, would be an asset;
• Professional fluency in English with excellent communication (both oral and written) skills; good command of at least one other language from the OSCE region is an asset;
• Previous professional experience in the OSCE area, in particular Central Asia or South-Eastern Europe, would be an asset;
• Knowledge of international human rights standards and ability to integrate these into own work, while complying with a do-no-harm approach;
• Proven didactic skills and ability to use modern, inclusive and interactive training approaches;
• Digital literacy and proficient use of Microsoft Office and relevant AI/ML demonstration tools;
• Flexibility and ability to work under time pressure to meet tight deadlines;
• Demonstrated gender awareness and sensitivity, and an ability to integrate a gender perspective into tasks and activities;
• Ability and willingness to work as a member of team, with people of different cultural, and religious backgrounds, different gender, and diverse political views, while maintaining impartiality and objectivity;

Remuneration Package:

Remuneration will be based on the selected consultant's/expert's qualifications, experience, the tasks and deliverables for this position and in accordance with the OSCE established rates.

The OSCE retains the discretion to re-advertise/re-post the vacancy, to cancel the recruitment or to offer an appointment with a modified job description or for a different duration.

Only those candidates who are selected to participate in the subsequent stages of recruitment will be contacted.

The OSCE is committed to diversity and inclusion within its workforce, and encourages qualified female and male candidates from all religious, ethnic and social backgrounds to apply to become a part of the Organization.

Candidates should be aware that OSCE officials shall conduct themselves at all times in a manner befitting the status of an international civil servant. This includes avoiding any action which may adversely reflect on the integrity, independence and impartiality of their position and function as officials of the OSCE. The OSCE is committed to applying the highest ethical standards in carrying out its mandate. For more information on the values set out in OSCE Competency Model,

Please be aware that the OSCE does not request payment at any stage of the application and review process.