Glemad

    Research programme 01

    GLEMAD RESEARCH

    Glemad Young Researchers Fellowship.

    Format
    Remote
    Track
    3 or 6 months
    Commitment
    5–10 hours each week

    ENTRY

    Research ability does not begin with an institutional title.

    The fellowship is for people at the beginning of serious work in AI security, infrastructure resilience, and accountable defensive systems. Selection is based on how an applicant thinks, communicates, and follows through.

    WHO SHOULD APPLY

    1. 01Students and self-taught researchers
    2. 02Engineers exploring AI security
    3. 03Analysts studying threats and infrastructure resilience
    4. 04Technical and analytical writers
    5. 05Early-career practitioners ready to complete independent research

    SEVEN FOCUS AREAS

    Choose one problem. Follow it far enough to produce something useful.

    Each area connects a research question to the conditions of live infrastructure: incomplete evidence, adversarial behavior, explicit authority, and real operational consequence.

    01

    Controlled autonomous defense

    Study how defensive authority can be scoped, constrained, and reviewed when a system can take action.

    02

    Safety-critical reasoning

    Examine reasoning where evidence is incomplete, decisions carry operational consequence, and stopping is a valid result.

    03

    Threat interpretation and adversarial simulation

    Develop ways to distinguish hostile behavior from legitimate change, then test those methods against adversarial scenarios.

    04

    Live infrastructure state and incident analysis

    Trace how events across identities, cloud services, runtime systems, and networks can form one changing incident.

    05

    Policy and compliance reasoning

    Explore how policy, evidence, and operating context can be represented inside a security decision.

    06

    Secure system behavior and evaluation

    Model and evaluate how security systems behave when inputs conflict, context changes, or limits are reached.

    07

    Applied research writing and technical documentation

    Turn methods, findings, limitations, and revisions into work another researcher can inspect.

    THE WORKING AGREEMENT

    A defined question, protected time, and a complete research output.

    Each stage narrows the work from a broad area to a question, a defensible method, and an inspectable contribution.

    1. 01

      Select a question

      Choose one focus area and define a research question that can be examined within the selected track.

    2. 02

      Work independently

      Study the problem, test ideas, document the method, and develop the work with a clear line of ownership.

    3. 03

      Challenge the work

      Test the method and claim against conflicting evidence, unresolved questions, and alternative explanations.

    4. 04

      Complete the record

      Finish one substantial output, such as a paper, technical report, evaluation, benchmark, prototype, simulation, or research brief.

    PROGRAMME SUPPORT

    What the fellowship provides, and when.

    Support follows the needs of the research. The conditions attached to publication, access, and recognition remain explicit.

    01
    PROGRAMME PRACTICE

    Research guidance

    Fellows meet Glemad researchers for monthly mentorship and receive asynchronous feedback as the work develops.

    02
    WORK-DEPENDENT

    A route to publication

    Fellows may publish papers, reports, benchmarks, or tools through the Glemad Research Index. Exceptional work may also be highlighted by Glemad.

    03
    PROJECT-DEPENDENT

    Supervised research access

    Relevant evaluation materials or research environments may be provided under supervision where a project requires them.

    04
    AT COMPLETION

    A formal programme record

    Fellows receive a certificate and public recognition at completion. An endorsement for academic or professional use is optional.

    SELECTION AND APPLICATION

    We assess the thinking, then the readiness to finish.

    Technical background can help, but it is not the deciding credential. The application should make your interest, reasoning, and available time clear.

    01

    How applications are assessed

    1. 01

      Clarity of interest in AI security

    2. 02

      Independent thought and clear communication

    3. 03

      Commitment to carry a project through

    4. 04

      Curiosity, discipline, and attention to detail

    5. 05

      Ethical awareness around safety-critical systems

    02

    Your application dossier

    • A short personal statement

      Required

    • A description of your research interests

      Required

    • Past code, writing, or project work

      Optional

    • Your preferred focus area

      Required

    • Availability for the selected track

      Required

    AFTER THE FELLOWSHIP

    A completed project can begin a longer exchange.

    RESEARCH COMMUNITY

    Fellows who complete the programme join the Glemad Research Community and can stay connected with other researchers.

    POSSIBLE CONTINUITY

    Future publication collaboration and continued guidance may follow where the work and research agenda align.

    BY INVITATION · NOT GUARANTEED

    Some fellows may be invited to continue a project or contribute to later research. Continuation is considered individually.

    ROLLING APPLICATIONS

    Show us the question you want to pursue.

    Tell us which focus area you want to investigate, why it matters to you, and how much time you can commit. Prior code, writing, or project work is optional.

    Apply now (opens in a new tab)Applications continue in the Glemad applicant portal.