Procurement workflow surface

Google DeepMind Frontier Safety Framework — vendor disclosure form

This is a sample disclosure form a procurement team can adapt for vendor RFPs and ITTs evaluating systems against Google DeepMind Frontier Safety Framework. The provision-specific questions below were derived from the catalog's coverage cells; before issuing, a qualified procurement lawyer should review the adapted version against your jurisdiction's contract law. This form is NOT legal advice (see charter §7.4).

Export:use your browser's Print → Save as PDF to produce a printable copy. Native .docx / .pdf export ships on the roadmap (see /wiki/roadmap); for now, browser print-to-PDF is the supported flow.

1. Vendor identification

Vendor legal name

Jurisdiction of incorporation

Primary contact (name, role)

Primary contact (email, phone)

RFP / ITT reference number

Response submission date

2. AI system identification

System name

Version / build identifier

Underlying model(s) + provider(s)

Intended use(s) within this engagement

Jurisdiction(s) of deployment

Other AI-governance instruments the system claims compliance with

3. Provision-specific questions

Foundation Models / GPAI. Does the offered system meet the threshold for a general-purpose / foundation model under Google DeepMind Frontier Safety Framework (FSF applies to Google DeepMind frontier-model releases)? If yes, identify the specific obligations you will satisfy and the evidence you will provide.
(Cite: FSF applies to Google DeepMind frontier-model releases)
Catastrophic & Existential Risk. Has the offered system been evaluated against catastrophic-risk thresholds (e.g., CBRN information uplift, autonomous replication) consistent with Google DeepMind Frontier Safety Framework (FSF Critical Capability Levels (CCL) — explicit thresholds for autonomy, biosecurity, cyber, persuasion)? Provide the evaluation report or its public-disclosure equivalent.
(Cite: FSF Critical Capability Levels (CCL) — explicit thresholds for autonomy, biosecurity, cyber, persuasion)
Agentic AI Governance. Does the offered system act as an autonomous agent (multi-step tool-use, transactions, recursion) within scope of Google DeepMind Frontier Safety Framework (FSF Critical Capability Levels — Autonomy is one of four named CCL domains)? Describe the tool-use surface, authorisation model, and human-oversight checkpoints.
(Cite: FSF Critical Capability Levels — Autonomy is one of four named CCL domains)

4. Documentation enclosures expected

Tick each enclosure attached to the vendor response. Missing enclosures should be explained in the “Variances” field below.

Safety / capability evaluation results
Vendor company registration + insurance certificates
Sub-processor / supply-chain list (including model upstream)

Variances (enclosures not attached + reason)

5. Vendor attestation

The undersigned, on behalf of the vendor, attests that the disclosures above are true and complete to the best of their knowledge at the date signed, and undertakes to notify the buyer in writing within 30 days of any material change to those disclosures.

Signatory name

Signatory role / title

Signature

Date

This is a sample form derived from the catalog at /wiki/deepmind-fsf. Adapt before issuing. Not legal advice; not jurisdiction-specific. See charter §7.4.

Google DeepMind Frontier Safety Framework — vendor disclosure form

3. Provision-specific questions

Foundation Models / GPAI. Does the offered system meet the threshold for a general-purpose / foundation model under Google DeepMind Frontier Safety Framework (FSF applies to Google DeepMind frontier-model releases)? If yes, identify the specific obligations you will satisfy and the evidence you will provide.

(Cite: FSF applies to Google DeepMind frontier-model releases)

Catastrophic & Existential Risk. Has the offered system been evaluated against catastrophic-risk thresholds (e.g., CBRN information uplift, autonomous replication) consistent with Google DeepMind Frontier Safety Framework (FSF Critical Capability Levels (CCL) — explicit thresholds for autonomy, biosecurity, cyber, persuasion)? Provide the evaluation report or its public-disclosure equivalent.

(Cite: FSF Critical Capability Levels (CCL) — explicit thresholds for autonomy, biosecurity, cyber, persuasion)

Agentic AI Governance. Does the offered system act as an autonomous agent (multi-step tool-use, transactions, recursion) within scope of Google DeepMind Frontier Safety Framework (FSF Critical Capability Levels — Autonomy is one of four named CCL domains)? Describe the tool-use surface, authorisation model, and human-oversight checkpoints.

(Cite: FSF Critical Capability Levels — Autonomy is one of four named CCL domains)