Organisation

NIST

NIST contributed to Digital Provenance and Authenticity Systems, Speech Recognition and Automated Transcription, Generative Language Models. NIST work on synthetic-content risks treats provenance, watermarking, detection and authentication as complementary approaches.

Type
governmental or standards institution
Known for
Digital Provenance and Authenticity Systems, Speech Recognition and Automated Transcription, Generative Language Models
Historical role

What the organisation contributed

NIST work on synthetic-content risks treats provenance, watermarking, detection and authentication as complementary approaches. NIST evaluation programmes made word error rate and standard corpora central to measurable progress. NIST's generative-AI profile later formalised risks including confabulation, data privacy, information integrity and harmful bias. It excludes ordinary photography, recording and animation; deterministic codecs; procedural graphics that follow explicit hand-authored rules without learned distributions; and conventional editing where all visible or audible source material was captured or manually created. It excludes a single generated answer with no external action; deterministic automation following a fully specified workflow without adaptive interpretation; ordinary recommendation systems; and the underlying language model considered independently. NIST characterises cloud computing through on-demand self-service, broad network access, resource pooling, rapid elasticity and measured service, with infrastructure, platform and software service models.

Why it mattered

Reduces uncertainty about a digital object’s claimed origin, custody and transformation history. Reduces the labour and delay required to convert speech into searchable symbolic text or commands.

Connected topics

Where this fits in ITEM

Evidence

Sources and further reading

  1. Artificial Intelligence Risk Management Framework

    NIST

    Open source ↗

  2. Artificial Intelligence Risk Management Framework (AI RMF 1.0)

    NIST · 2023

    Open source ↗

  3. Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile

    NIST · 2024

    Open source ↗

  4. ASCII reference

    NIST

    Open source ↗

  5. Binary made many kinds of information look alike to machinery. Standards and software taught the machinery what the patterns were supposed to mean

    Open source ↗

  6. Digital Signature Standard:

    NIST

    Open source ↗

  7. OpenASR20: An Open Challenge for Automatic Speech Recognition of Conversational Telephone Speech

    NIST

    Open source ↗

  8. OpenASR21: The Second Open Challenge for Automatic Speech Recognition for Low-Resource Languages

    NIST

    Open source ↗

  9. OpenSAT 2020 Evaluation Plan

    NIST · 2020

    Open source ↗

  10. Reducing Risks Posed by Synthetic Content

    NIST · 2024

    Open source ↗

  11. Tallies remembered that something happened. Numerical notation made quantity available for thought, audit and power

    Open source ↗

  12. The TREC Spoken Document Retrieval Track: A Success Story

    NIST

    Open source ↗