I will direct NIST to develop a structured, consensus-based AI model evaluation documentation template, with modular sections, technical guidance, and a pilot tool for voluntary use.

The READ AI Models Act would: Direct NIST to develop a structured, consensus-based template for AI model evaluation documentation, similar to a “nutrition label” that developers can voluntarily use. Require that the template be modular, allowing organizations to adopt only the sections relevant to their sector or use case. Produce technical guidance that incorporates voluntary standards, benchmarks, and industry best practices. Launch a pilot tool to help users generate documentation easily and consistently. Require NIST to seek public comment, collaborate with researchers, industry, and international standards bodies, and publish the final template and guidance publicly.

McBride-backed legislation would direct NIST to create a voluntary AI documentation template, guidance, and pilot tool.

McBride Introduces Bipartisan “READ AI Models Act” to Bring More Transparency to Artificial Intelligence | Congresswoman Sarah Mcbride
primary · press_release · model gpt-5.4-mini

A June 2 executive order created a voluntary federal review framework for AI models, asking companies to share models with the government up to 30 days before public release and stopping short of mandatory preclearance.

This is adjacent progress toward voluntary AI evaluation, but it does not show the promised NIST documentation template, modular guidance, or pilot tool was produced.

partial same_term

Trump signs executive order seeking early access to new AI releases | Donald Trump | The Guardian
secondary · model gpt-5.4-mini · confidence 72%

On June 9, reports said the White House told CAISI to halt public model-assessment reports, kept internal evaluations going, and removed a recent announcement about collaborations with major tech firms.

This looks like a constraint or reversal on the public-facing evaluation effort rather than delivery of the requested NIST template and voluntary pilot tool.

never same_term

White House Reins In AI-Testing Unit as National-Security Concerns Grow
secondary · model gpt-5.4-mini · confidence 81%

Rep. McBride's official House page directs constituents to legislation she sponsored and co-sponsored and lists recent votes; the visible recent votes concern unrelated bills such as Taiwan, housing, D.C. taxation, critical minerals, appropriations, and veterans benefits.

The member's own legislation page does not show a McBride-sponsored NIST AI model evaluation documentation template or related pilot-tool action in the visible record.

never same_term

Votes and Legislation | Congresswoman Sarah McBride
secondary · model gpt-5.5 · confidence 78%

NIST describes the AI RMF as voluntary, developed through a consensus-driven public process, and lists the AI RMF 1.0, Playbook, Resource Center, Generative AI Profile, and a 2026 critical-infrastructure profile concept note.

NIST has broad voluntary AI risk-management materials, but this page does not document the promised structured AI model evaluation documentation template with modular sections and a pilot tool.

partial unknown

AI Risk Management Framework | NIST
secondary · model gpt-5.5 · confidence 84%

NIST says the AI RMF Playbook is a voluntary companion with suggested actions, references, and guidance for the Govern, Map, Measure, and Manage functions; it was first completed in March 2023 and will be updated after AI RMF 1.0 is revised.

The Playbook is voluntary technical guidance, but it predates McBride's House term and is not the specific model-evaluation documentation template or pilot tool promised.

partial unknown

NIST AI RMF Playbook | NIST
secondary · model gpt-5.5 · confidence 82%

NIST says CAISI will develop guidelines and best practices, assist industry with voluntary standards, establish voluntary agreements with AI developers and evaluators, and lead unclassified evaluations of AI capabilities that may pose national-security risks.

CAISI activity is adjacent to AI model evaluation and voluntary standards, but the page does not show McBride-directed delivery of a modular documentation template or voluntary pilot tool.

partial same_term

Center for AI Standards and Innovation (CAISI) | NIST
secondary · model gpt-5.5 · confidence 83%

NIST announced a CAISI initial public draft on practices for automated benchmark evaluations of language models, with sections on defining objectives, implementing and running evaluations, and analyzing and reporting results; comments were due March 31, 2026.

This is the closest official NIST work found: it advances evaluation-reporting guidance, but it is a draft best-practices document, not the promised consensus-based documentation template with modular sections and pilot tool, and it is not tied to McBride action.

partial same_term

Towards Best Practices for Automated Benchmark Evaluations | NIST
secondary · model gpt-5.5 · confidence 87%

The White House action plan recommends that NIST and CAISI publish guidelines and resources for federal agencies to evaluate AI systems, support the science of AI model measurement, convene best-practice meetings, and invest in testbeds for piloting AI systems.

The plan contains adjacent executive-branch direction for AI evaluation resources and testbeds, but it does not establish McBride credit or the specific NIST documentation template and voluntary pilot tool described in the promise.

partial same_term

America's AI Action Plan
secondary · model gpt-5.5 · confidence 80%

NIST says CAISI and GSA agreed to support USAi evaluation needs, develop methodologies for evaluating performance, security, and functionality in federal workflows, and create resources including pre-deployment assessment guidelines and tools for post-deployment performance measurement.

This shows official tools and guidance for federal AI procurement evaluation, but it is not a public voluntary AI model evaluation documentation template and no McBride role is shown.

partial same_term

CAISI signs MOU with GSA to boost AI evaluation science in federal procurement through USAi | NIST
secondary · model gpt-5.5 · confidence 79%

NIST describes CAISI's report on post-deployment monitoring as identifying monitoring categories, challenges, gaps, barriers, and open questions; highlighted gaps include a lack of trusted guidelines or standards for methods and tools.

The report underscores that NIST was still mapping gaps and open questions in AI monitoring rather than delivering the promised evaluation documentation template and pilot tool.

never same_term

New Report: Challenges to the Monitoring of Deployed AI Systems | NIST
secondary · model gpt-5.5 · confidence 84%

CAISI published an evaluation of DeepSeek V4 Pro using capability benchmarks across cyber, software engineering, natural sciences, abstract reasoning, and mathematics, including some non-public CAISI benchmarks.

This proves NIST/CAISI performed model evaluations, but it is a one-model assessment and benchmark report, not the promised reusable documentation template or voluntary pilot tool, and it is not attributable to McBride.

partial same_term

CAISI Evaluation of DeepSeek V4 Pro | NIST
secondary · model gpt-5.5 · confidence 76%

The Wall Street Journal reported that administration officials asked CAISI to stop publishing public model-assessment reports while a new executive order was implemented, while internal model evaluation work continued.

Reported restrictions on CAISI's public-facing model assessment work weigh against finding that the promised public voluntary documentation template and pilot tool had been delivered by the assessment date.

never same_term

White House Reins In AI-Testing Unit as National-Security Concerns Grow
secondary · model gpt-5.5 · confidence 72%