IARPA DECIPHER

IARPA DECIPHER is the ETA topic for multilingual specialised language detection. It is designed to detect and define acronyms, jargon, slang, coded language, and other expressions that are difficult to disambiguate across diverse text collections, especially where standard translation or NLP tools fall short. The programme focuses on language analysis for intelligence use, not on general translation quality. DECIPHER follows the same Prototype Other Transaction Agreement route under DOI-ETA-FY26-30. White papers are accepted on a rolling basis after the Q&A posting, the initial wave was due 30 June 2026, and no public award amount is given in the available sources. The prime contractor must be US-based, while universities, nonprofits, small businesses, and other eligible teams can participate. Government agencies, FFRDCs, and UARCs are excluded, and non-US organisations may participate only as subcontractors to US teams. The best fit is a team that can handle multilingual ambiguity with discipline and show a path from language detection to an intelligence workflow. Because the topic is built around coded and specialised phrasing, the strongest proposals will keep the scope narrow, use concrete text-analysis examples, and demonstrate that the system can cope with terminology that shifts meaning across contexts. The programme rewards technical precision more than broad coverage.

Each grant below is a distinct funding opportunity with its own eligibility, scope, and deliverables.