Mozilla Technology Fund — AI Auditing Tools

Criteria breakdown

Red flags

• Core topic (AI auditing tools for algorithmic accountability) sits outside ETC's humanitarian intervention wheelhouse—fit requires reframing a project around a human-rights audit use case.
• Mandatory open-source output may conflict with partner/agency deployment models ETC often uses.
• Small award ($50K) relative to the strategic effort of entering a competitive, specialized applicant field.
• Only worth pursuing if ETC has a concrete audit-tool concept tied to migration, forced labor, or traceability ready to propose.

Tips, most important first

• highPositioningReframe Traceability as an auditor-facing tool, not a supply-chain platform

  MTF funds tools FOR auditors creating accountability, not intervention systems. Recast your supply-chain traceability work as a benchmark/dataset or analysis tool that lets labor-rights auditors verify or falsify a company's ethical-sourcing claims — a concrete, reusable audit artifact in the spirit of the 2023 'observatory' and pricing-inspection tools.

• highPitfall to avoidResolve the open-source conflict before you write a word

  Every line of code must ship open source and content under Creative Commons — this collides with the closed partner/agency deployment models ETC often uses. Confirm with your intended partner (and Microsoft Research collaborators) that the specific audit tool can be fully OSS, or the proposal is dead on arrival.

• highFunder researchAnchor explicitly to the OAT Initiative and its people

  Winners must contribute to the Open Source Audit Tooling Initiative led by Deb Raji (with Birhane, Vecchione, Steed, Ojewale). Read OAT's published standards and name how your tool plugs into them — reviewers reward interoperability with the existing community, not a standalone artifact.

• highPositioningMatch the $50K scope: one shippable tool, not a research agenda

  The 2022–2023 cohorts funded tightly-scoped, shippable tools (crowdsourced voice-assistant bias measurement, a content-promotion observatory). Scope your deliverable to something a small team can build and release in 12 months, and lead with the specific accountability decision it enables for a named auditor type.

• highProposal craftLead with a working prototype, not a concept

  MTF applicants 'generally need a working prototype.' Show a demonstrable slice of your traceability tooling now and evidence your interdisciplinary team (policy + technical prototyping, CGA + Microsoft Research) can execute over 12 months — this is where ETC's applied-prototyping strength is your real differentiator over pure-tech applicants.

• mediumRelationshipUse the pre-deadline window to test fit with Mozilla/OAT

  Applications open ~Sept 6 for an Oct 5 close. Before then, engage the OAT community (public channels, MozFest presence, Deb Raji's published work) to sanity-check whether a forced-labor/traceability audit tool is welcome — this de-risks a low-probability bet and signals you're a genuine community contributor.

• mediumProcess & logisticsPlan for the two-stage Fluxx flow and cohort commitment

  This is LOI → invited full application via Fluxx, with monthly virtual convenings and mandatory MozFest attendance. Confirm a named team member can carry the cohort obligations for a full year before committing, and map backward from the Oct 5 LOI so the invited full app isn't rushed.

• mediumBudgetFrame the $50K around researcher/build time, honestly

  The award is small and NYU overhead could swallow it — structure the budget toward developer/researcher time to build and release the open tool, and confirm ETC's constraint of funding staff time over pure overhead can be honored within Mozilla's allowable costs.

• mediumPitfall to avoidBe willing to defer if no concrete tool concept is ready

  With fit at 49/100 and win probability ~15%, this is only worth it if a specific migration/forced-labor/traceability audit-tool concept is already crystallized. If it isn't by early September, skip this cycle — MTF runs annually, so a stronger prototype next year beats a thin one now.

Next actions

1. 1Convene the Traceability team to decide if a single, OSS-releasable auditor-facing tool (e.g. a forced-labor claims-verification benchmark or analysis tool) can be defined and prototyped — go/no-go on pursuing this cycle.
2. 2Get written confirmation from the relevant partner/agency and Microsoft Research that the specific tool can be fully open source under CC before drafting.
3. 3Read the OAT Initiative's published standards and 2022–2023 MTF winners in detail, and draft one paragraph on how ETC's tool interoperates with OAT.
4. 4Engage the OAT/Mozilla community informally (or via MozFest channels) to test-fit the forced-labor/traceability audit concept before the September opening.
5. 5Assemble prototype evidence and name the team member who will own the 12-month cohort commitment, then map an internal timeline backward from the Oct 5 LOI.
6. 6Draft a tight, single-tool LOI scoped to the $50K/12-month envelope, with a budget weighted toward build/research time.