private company / nonprofit-controlled structure

OpenAI

An AI company with a nonprofit-controlled corporate structure selling models, APIs, and ChatGPT, with concentrated governance, safety, labor, and platform-power questions. OpenAI remains weak because of person-dependent governance, documented human harm, and ecological harm, even where the product or service has practical value.

Why this matters: AI infrastructure can become civic infrastructure before accountability structures mature.

Letter grade D Extractive Medium confidence (AI) Rubric gcd-rubric-v1 Last reviewed May 6, 2026

Final Score 5* D - Extractive

DEXTRACTIVE5 pts

* Tentative scaffolding score. Not human-checked or final.

Base Material27

Bonus+0

Cap41Ownership <= 4 and Governance <= 4

After Cap27

Penalties-22

Private Government badge

OpenAI receives the Private Government badge because it combines large-scale private rule-setting over customers, workers, users, members, or counterparties with OpenLobby-documented lobbying of about $680K on public policy issues.

Evidence: [16]

!Not Verified

Represent this organization?

Request verification to have the evidence record checked against primary materials and organization-supplied documents. Verification does not buy a higher score; scores change only when review finds factual errors, omissions, or miscalibration.

Request verification

Evaluation Overview

D - Extractive

The score turns mainly on Extraction, with the largest penalty coming from Reversibility.

Strengths

Extraction4/10
Ownership3/10
Governance3/10

Penalties

Reversibility-5
Policy Capture-5
Human Harm-4
Ecological Harm-4

Evidence state

ConfidenceMedium confidence (AI)
ThoroughnessDeveloped (AI)
Linked claims16
Direct axis claims15
Coverage15/15

Scoring Axes

Axis	Score	Why this score
Ownership ? Control rights: shareholder-dominated at 0, worker cooperative control at 10.	3 / 10	OpenAI describes a for-profit business structure governed and controlled by the nonprofit OpenAI Foundation, with ongoing recapitalization into a public benefit corporation structure. OpenAI describes a for-profit business structure governed and controlled by the nonprofit OpenAI Foundation, with ongoing recapitalization into a public benefit corporation structure. That keeps OpenAI low on Ownership: decisive control remains with shareholders, private owners, executives, or state insiders rather than affected people. The value is above zero only where the linked record still shows some real function, constraint, or public value on this axis. Calibration notes Comparative anchor: Column-spectrum position: 100 scored entries below, 9 tied, 117 above on Ownership. Linked evidence context OpenAI describes a for-profit business structure governed and controlled by the nonprofit OpenAI Foundation, with ongoing recapitalization into a public benefit corporation structure. [1] OpenAI describes a for-profit business structure governed and controlled by the nonprofit OpenAI Foundation, with a complex structure that is not direct user, worker, or public ownership. [2]
Governance ? Binding decision authority: centralized control at 0, democratic stakeholder control at 10.	3 / 10	OpenAI describes a for-profit business structure governed and controlled by the nonprofit OpenAI Foundation, with ongoing recapitalization into a public benefit corporation structure. OpenAI describes a for-profit business structure governed and controlled by the nonprofit OpenAI Foundation, with ongoing recapitalization into a public benefit corporation structure. That keeps OpenAI low on Governance: ordinary users, workers, customers, or affected communities cannot change the rules that govern them. The value is above zero only where the linked record still shows some real function, constraint, or public value on this axis. Calibration notes Comparative anchor: Column-spectrum position: 94 scored entries below, 13 tied, 119 above on Governance. Linked evidence OpenAI describes a for-profit business structure governed and controlled by the nonprofit OpenAI Foundation, with ongoing recapitalization into a public benefit corporation structure. [1]
Extraction ? Surplus allocation, wage share, CEO pay ratio, margins, and structured extraction judgment.	4 / 10	OpenAI describes a for-profit business structure governed and controlled by the nonprofit OpenAI Foundation, with ongoing recapitalization into a public benefit corporation structure. OpenAI describes a for-profit business structure governed and controlled by the nonprofit OpenAI Foundation, with ongoing recapitalization into a public benefit corporation structure. That places OpenAI in the middle of Extraction: the institution creates real value but still leaves room for rent extraction, unequal bargaining, or surplus capture. The value stays below the top tier because the linked record does not show full democratic control, non-extractive lock-in, or complete stakeholder protection on this axis. Calibration notes Comparative anchor: Column-spectrum position: 62 scored entries below, 32 tied, 132 above on Extraction. Linked evidence context OpenAI describes a for-profit business structure governed and controlled by the nonprofit OpenAI Foundation, with ongoing recapitalization into a public benefit corporation structure. [1] OpenAI's structure page and Microsoft partnership materials describe a commercial AI business dependent on large cloud partnership and product monetization rather than a public utility model. [3]
Labor Sovereignty ? Worker power: coercive conditions at 0, co-determination or ownership at 7. Employee dissatisfaction matters only when source-backed evidence shows concrete limits on worker agency, such as coercive scheduling, retaliation, wage theft, harassment, unsafe conditions, suppression of worker voice, or extreme turnover.	3 / 7	OpenAI describes a for-profit business structure governed and controlled by the nonprofit OpenAI Foundation, with ongoing recapitalization into a public benefit corporation structure. OpenAI describes a for-profit business structure governed and controlled by the nonprofit OpenAI Foundation, with ongoing recapitalization into a public benefit corporation structure. That places OpenAI in the middle of Labor Sovereignty: workers or participants have some protection or agency, but not enough binding power to control the relationship. The value stays below the top tier because the linked record does not show full democratic control, non-extractive lock-in, or complete stakeholder protection on this axis. Calibration notes Comparative anchor: Column-spectrum position: 45 scored entries below, 66 tied, 115 above on Labor Sovereignty. Linked evidence context OpenAI describes a for-profit business structure governed and controlled by the nonprofit OpenAI Foundation, with ongoing recapitalization into a public benefit corporation structure. [1] OpenAI's public structure gives employees no disclosed worker-ownership or democratic labor-control mechanism over model deployment, commercialization, or governance tradeoffs. [4]
Solidarity with the Unemployed ? Treatment of exits and nonworkers, including severance, redeployment, and non-competes.	3 / 7	OpenAI describes a for-profit business structure governed and controlled by the nonprofit OpenAI Foundation, with ongoing recapitalization into a public benefit corporation structure. OpenAI describes a for-profit business structure governed and controlled by the nonprofit OpenAI Foundation, with ongoing recapitalization into a public benefit corporation structure. That places OpenAI in the middle of Solidarity with the Unemployed: some access, severance, retraining, openness, or mutual-aid value is present, but it is not central or durable. The value stays below the top tier because the linked record does not show full democratic control, non-extractive lock-in, or complete stakeholder protection on this axis. Calibration notes Comparative anchor: Column-spectrum position: 81 scored entries below, 45 tied, 100 above on Solidarity with the Unemployed. Linked evidence context OpenAI describes a for-profit business structure governed and controlled by the nonprofit OpenAI Foundation, with ongoing recapitalization into a public benefit corporation structure. [1] OpenAI's charter states a mission to benefit humanity, but the reviewed public structure does not guarantee access, bargaining power, or protection for people displaced or governed by AI systems. [5]
Loss-Bearing Fidelity ? Willingness to absorb costs to preserve values, workers, users, and public obligations.	3 / 7	OpenAI describes a for-profit business structure governed and controlled by the nonprofit OpenAI Foundation, with ongoing recapitalization into a public benefit corporation structure. OpenAI describes a for-profit business structure governed and controlled by the nonprofit OpenAI Foundation, with ongoing recapitalization into a public benefit corporation structure. That places OpenAI in the middle of Loss-Bearing Fidelity: there is some willingness to preserve obligations, but the record still leaves meaningful risk-shifting concerns. The value stays below the top tier because the linked record does not show full democratic control, non-extractive lock-in, or complete stakeholder protection on this axis. Calibration notes Comparative anchor: Column-spectrum position: 48 scored entries below, 39 tied, 139 above on Loss-Bearing Fidelity. Linked evidence context OpenAI describes a for-profit business structure governed and controlled by the nonprofit OpenAI Foundation, with ongoing recapitalization into a public benefit corporation structure. [1] OpenAI's charter acknowledges broad AGI stakes, while its commercial deployment structure leaves many model-risk, labor, education, misinformation, and environmental costs outside direct user control. [6]
Market Conduct ? Pricing fairness, switching costs, lock-in, and rent extraction.	2 / 5	OpenAI and Microsoft disclose a strategic partnership and commercial cloud relationship around OpenAI models and infrastructure. OpenAI and Microsoft disclose a strategic partnership and commercial cloud relationship around OpenAI models and infrastructure. That places OpenAI in the middle of Market Conduct: the product has real value but still uses meaningful lock-in, gatekeeping, market power, or dependency. The value stays below the top tier because the linked record does not show full democratic control, non-extractive lock-in, or complete stakeholder protection on this axis. Calibration notes Comparative anchor: Column-spectrum position: 31 scored entries below, 22 tied, 173 above on Market Conduct. Linked evidence OpenAI and Microsoft disclose a strategic partnership and commercial cloud relationship around OpenAI models and infrastructure. [7]
Product Integrity ? Preservation of quality rather than degradation for monetization.	3 / 5	OpenAI publishes a charter stating that its mission is to ensure artificial general intelligence benefits all of humanity. OpenAI publishes a charter stating that its mission is to ensure artificial general intelligence benefits all of humanity. That places OpenAI in the middle of Product Integrity: the product has real use value but is compromised by incentives, monetization, lock-in, quality degradation, or risk. The value stays below the top tier because the linked record does not show full democratic control, non-extractive lock-in, or complete stakeholder protection on this axis. Calibration notes Comparative anchor: Column-spectrum position: 37 scored entries below, 23 tied, 166 above on Product Integrity. Linked evidence OpenAI publishes a charter stating that its mission is to ensure artificial general intelligence benefits all of humanity. [8]
Scale Integrity ? Whether growth improves or degrades fairness and accountability.	3 / 5	OpenAI and Microsoft disclose a strategic partnership and commercial cloud relationship around OpenAI models and infrastructure. OpenAI and Microsoft disclose a strategic partnership and commercial cloud relationship around OpenAI models and infrastructure. That places OpenAI in the middle of Scale Integrity: scale brings useful reach but also weakens accountability, transparency, or local constraint. The value stays below the top tier because the linked record does not show full democratic control, non-extractive lock-in, or complete stakeholder protection on this axis. Calibration notes Comparative anchor: Column-spectrum position: 50 scored entries below, 58 tied, 118 above on Scale Integrity. Linked evidence context OpenAI and Microsoft disclose a strategic partnership and commercial cloud relationship around OpenAI models and infrastructure. [7] OpenAI's frontier-AI scale and cloud partnership make its systems broadly consequential while remaining controlled through private and nonprofit corporate governance rather than democratic public infrastructure. [9]

Penalties

Penalty	Applied	Why this penalty
Reversibility ? Applied when positive conduct depends on current living leaders, founders, family owners, or other person-contingent governance rather than durable structure. Range: -5 to 0.	-5	OpenAI's favorable conduct depends partly on current leaders, founders, family owners, executive discretion, or other person-contingent control rather than a fully locked democratic stakeholder-control structure. OpenAI's favorable conduct depends partly on current leaders, founders, family owners, executive discretion, or other person-contingent control rather than a fully locked democratic stakeholder-control structure. That warrants a severe Reversibility penalty because favorable conduct depends on current people or discretionary leadership rather than locked stakeholder structure. The value is near the floor because the conduct appears central, large-scale, repeated, or institutionally powerful rather than incidental. Calibration notes Comparative anchor: Column-spectrum position: 0 scored entries below, 95 tied, 131 above on Reversibility. Linked evidence context OpenAI describes a for-profit business structure governed and controlled by the nonprofit OpenAI Foundation, with ongoing recapitalization into a public benefit corporation structure. [1] OpenAI's favorable conduct depends partly on current leaders, founders, family owners, executive discretion, or other person-contingent control rather than a fully locked democratic stakeholder-control structure. [10]
Human Harm ? Severity, scale, culpability, vulnerable targets, willful reoffense, reputation laundering, pattern escalation, and time decay. Range: -20 to 0.	-4	Multiple lawsuits reported in 2025-2026 allege severe self-harm or violence linked to ChatGPT interactions; these remain allegations, so the penalty is material but below adjudicated severe-harm anchors. Multiple lawsuits reported in 2025-2026 allege severe self-harm or violence linked to ChatGPT interactions; these remain allegations, so the penalty is material but below adjudicated severe-harm anchors. That warrants a limited Human Harm penalty because ordinary operation, documented misconduct, or institutional power creates serious injury, deprivation, abuse, or foreseeable human damage. The value is limited because the linked record shows a real problem but not one that dominates the profile on this axis. Calibration notes Comparative anchor: Column-spectrum position: 48 scored entries below, 3 tied, 175 above on Human Harm. Linked evidence Multiple lawsuits reported in 2025-2026 allege severe self-harm or violence linked to ChatGPT interactions; these remain allegations, so the penalty is material but below adjudicated severe-harm anchors. [11]
Ecological Harm ? Footprint, trajectory, irreversibility, culpability, and scale, including whether ecological harm is central to the model and continued after credible notice. Range: -30 to 0.	-4	OpenAI's products depend on large-scale AI compute; OpenAI Academy materials discuss environmental impacts of AI, including energy and efficiency questions, describes conduct considered under Ecological Harm. OpenAI's products depend on large-scale AI compute; OpenAI Academy materials discuss environmental impacts of AI, including energy and efficiency questions, describes conduct considered under Ecological Harm. That warrants a limited Ecological Harm penalty because the institution causes or defends ecological damage that is more than incidental to the model. The value is limited because the linked record shows a real problem but not one that dominates the profile on this axis. Calibration notes Comparative anchor: Column-spectrum position: 33 scored entries below, 9 tied, 184 above on Ecological Harm. Linked evidence OpenAI's products depend on large-scale AI compute; OpenAI Academy materials discuss environmental impacts of AI, including energy and efficiency questions, describes conduct considered under Ecological Harm. [12]
Policy Capture ? Private-interest attempts to bend law, regulation, subsidies, taxes, enforcement, trade, labor, safety, competition, environmental, healthcare, housing, surveillance, civil-liberties, or consumer policy against workers, customers, citizens, affected communities, or ecological life. Public-interest advocacy is not penalized merely because it is lobbying. Range: -15 to 0.	-5	OpenLobby records OPENAI lobbying on technology, science / technology, consumer product safety with about $680K in spending across 6 filings; the reviewed filings document private… OpenLobby records OPENAI lobbying on technology, science / technology, consumer product safety with about $680K in spending across 6 filings; the reviewed filings document private institutional lobbying over public rules rather than a public-interest advocacy role. That warrants a limited Policy Capture penalty because private institutional pressure is being applied to public rules while affected people cannot democratically govern that agenda. The value is limited because the linked record shows a real problem but not one that dominates the profile on this axis. Linked evidence OpenLobby records OPENAI lobbying on technology, science / technology, consumer product safety with about $680K in spending across 6 filings; the reviewed filings document private institutional lobbying over public rules rather than a public-interest advocacy role. [15]
Accountability Opacity ? Material opacity, reputation laundering, or hidden accountability structures that prevent public accountability. Range: -2 to 0.	-2	OpenAI's nonprofit-control and restructuring debate creates a accountability-opacity penalty because public accountability depends on understanding a complex nonprofit/for-profit control structure under regulatory and civic pressure. OpenAI's nonprofit-control and restructuring debate creates a accountability-opacity penalty because public accountability depends on understanding a complex nonprofit/for-profit control structure under regulatory and civic pressure. That warrants a severe Accountability Opacity penalty because material facts, risks, conflicts, ownership, or accountability information are obscured from the public. The value is near the floor because the conduct appears central, large-scale, repeated, or institutionally powerful rather than incidental. Calibration notes Comparative anchor: Column-spectrum position: 0 scored entries below, 52 tied, 174 above on Accountability Opacity. Linked evidence OpenAI's nonprofit-control and restructuring debate creates a accountability-opacity penalty because public accountability depends on understanding a complex nonprofit/for-profit control structure under regulatory and civic pressure. [13]
Ideological Disavowal ? Concealed ideology presented as neutrality, expertise, professional necessity, public-service administration, market inevitability, or non-ideological common sense while exercising power. Range: -3 to 0.	-2	OpenAI's public framing is assessed as ideological-disavowal evidence because OpenAI exercises meaningful institutional or cultural power while strongly presenting contested choices as… OpenAI's public framing is assessed as ideological-disavowal evidence because OpenAI exercises meaningful institutional or cultural power while strongly presenting contested choices as neutral expertise, safety, objectivity, professionalism, administrative necessity, or public service. That warrants a substantial Ideological Disavowal penalty because the institution presents situated, contested choices as neutrality, expertise, inevitability, safety, or common sense. The value is substantial but not at the floor because the linked record shows serious conduct without the strongest aggravating factors on this axis. Calibration notes Comparative anchor: Ideological Disavowal 0 to -3 recalibration: 0 for explicit standpoint or no penalty-level neutrality theater; -1 for bounded softening of contested choices; -2 for strong neutrality theater by a powerful institution; -3 for major gatekeeping or coercive power presented as neutral, necessary, objective, or non-ideological. Uncertainty This recalibration uses public self-presentation and profile evidence; later review should add direct language-analysis evidence for high-impact institutions. Linked evidence OpenAI's public framing is assessed as ideological-disavowal evidence because OpenAI exercises meaningful institutional or cultural power while strongly presenting contested choices as neutral expertise, safety, objectivity, professionalism, administrative necessity, or public service. [14]

Bonus Credits

No bonus credits applied in this version.

Confidence Basis

Confidence is computed from the evidence trail and review state, not typed into the profile by hand.

Medium 79/100

15 verified linked claims 15 direct axis claims 0 disputed claims 15/15 components covered

This confidence label measures the source-backed evidence trail. AI-scaffolded scores remain tentative until human review.

Claim confidence 15/20

15 verified linked claims

Source quality 11/18

Best source per verified claim, weighted by institutional reliability

Direct axis-specific claims 14/18

15 direct claims across 15 active components

Dispute load 12/12

0 disputed claims on this entity

Recency 10/10

Newest accepted timestamp: May 13, 2026

Reviewer status 7/12

Human-reviewed components score higher than AI scaffolding

Component coverage 10/10

15/15 evidence-bearing components have direct support

Evidence State

Profile stateAI draft / human-pending

VerificationUnverified

ConfidenceMedium confidence (AI)

ThoroughnessDeveloped (AI)

Correction routeUse “Challenge this rating” for factual errors, missing counterevidence, source problems, or calculation mistakes.

Company responseCompany representatives can submit source-backed corrections; payment never changes scores or reviewer authority.

Claims and Sources

Claims are the evidence record. Each claim needs a source link, axis category, status, confidence level, and timestamp before it can support a score.

* Tentative scaffolding score. Not human-checked or final.

1OpenAI describes a for-profit business structure governed and controlled by the nonprofit OpenAI Foundation, with ongoing recapitalization into a public benefit corporation structure.

Governance Verified High confidence AI-generated / human-pending May 5, 2026

AI scaffold

OpenAI - Our structureGeneral web source · 58%

2OpenAI describes a for-profit business structure governed and controlled by the nonprofit OpenAI Foundation, with a complex structure that is not direct user, worker, or public ownership.

Ownership Verified High confidence AI-generated / human-pending May 13, 2026

AI scaffold

OpenAI structureGeneral web source · 58%

3OpenAI's structure page and Microsoft partnership materials describe a commercial AI business dependent on large cloud partnership and product monetization rather than a public utility model.

Extraction Verified High confidence AI-generated / human-pending May 13, 2026

AI scaffold

OpenAI structureGeneral web source · 58%
Microsoft and OpenAI extend partnershipGeneral web source · 58%

4OpenAI's public structure gives employees no disclosed worker-ownership or democratic labor-control mechanism over model deployment, commercialization, or governance tradeoffs.

Labor Sovereignty Verified Medium confidence AI-generated / human-pending May 13, 2026

AI scaffold

OpenAI structureGeneral web source · 58%

5OpenAI's charter states a mission to benefit humanity, but the reviewed public structure does not guarantee access, bargaining power, or protection for people displaced or governed by AI systems.

Solidarity Unemployed Verified Medium confidence AI-generated / human-pending May 13, 2026

AI scaffold

OpenAI CharterGeneral web source · 58%
OpenAI structureGeneral web source · 58%

6OpenAI's charter acknowledges broad AGI stakes, while its commercial deployment structure leaves many model-risk, labor, education, misinformation, and environmental costs outside direct user control.

Loss Bearing Fidelity Verified Medium confidence AI-generated / human-pending May 13, 2026

AI scaffold

OpenAI CharterGeneral web source · 58%
OpenAI structureGeneral web source · 58%

7OpenAI and Microsoft disclose a strategic partnership and commercial cloud relationship around OpenAI models and infrastructure.

Market Conduct Verified Medium confidence AI-generated / human-pending May 5, 2026

AI scaffold

Microsoft and OpenAI extend partnershipGeneral web source · 58%

8OpenAI publishes a charter stating that its mission is to ensure artificial general intelligence benefits all of humanity.

Product Integrity Verified High confidence AI-generated / human-pending May 5, 2026

AI scaffold

OpenAI CharterGeneral web source · 58%

9OpenAI's frontier-AI scale and cloud partnership make its systems broadly consequential while remaining controlled through private and nonprofit corporate governance rather than democratic public infrastructure.

Scale Integrity Verified High confidence AI-generated / human-pending May 13, 2026

AI scaffold

OpenAI structureGeneral web source · 58%
Microsoft and OpenAI extend partnershipGeneral web source · 58%

10OpenAI's favorable conduct depends partly on current leaders, founders, family owners, executive discretion, or other person-contingent control rather than a fully locked democratic stakeholder-control structure.

Reversibility Verified Medium confidence Human-reviewed May 5, 2026

OpenAI - Our structureGeneral web source · 58%

11Multiple lawsuits reported in 2025-2026 allege severe self-harm or violence linked to ChatGPT interactions; these remain allegations, so the penalty is material but below adjudicated severe-harm anchors.

Human Harm Verified Medium confidence Human-reviewed May 5, 2026

AP: OpenAI faces lawsuits claiming ChatGPT drove people to suicide, delusionsIndependent reporting · 82%
CNBC: OpenAI plans ChatGPT changes after lawsuitGeneral web source · 58%

12OpenAI's products depend on large-scale AI compute; OpenAI Academy materials discuss environmental impacts of AI, including energy and efficiency questions, describes conduct considered under Ecological Harm.

Ecological Harm Verified Medium confidence Human-reviewed May 5, 2026

OpenAI Academy: Environmental Impact of AIGeneral web source · 58%

13OpenAI's nonprofit-control and restructuring debate creates a accountability-opacity penalty because public accountability depends on understanding a complex nonprofit/for-profit control structure under regulatory and civic pressure.

Accountability Opacity Verified Medium confidence Human-reviewed May 5, 2026

OpenAI - Our structureGeneral web source · 58%
CNBC: OpenAI says nonprofit will retain controlGeneral web source · 58%

14OpenAI's public framing is assessed as ideological-disavowal evidence because OpenAI exercises meaningful institutional or cultural power while strongly presenting contested choices as neutral expertise, safety, objectivity, professionalism, administrative necessity, or public service.

Ideological Disavowal Verified Medium confidence AI-generated / human-pending May 9, 2026

AI scaffold

OpenAI official siteGeneral web source · 58%
OpenAI CharterGeneral web source · 58%
Microsoft and OpenAI extend partnershipGeneral web source · 58%

15OpenLobby records OPENAI lobbying on technology, science / technology, consumer product safety with about $680K in spending across 6 filings; the reviewed filings document private institutional lobbying over public rules rather than a public-interest advocacy role.

Policy Capture Verified Medium confidence AI-generated / human-pending May 11, 2026

AI scaffold

OpenLobby — OPENAI lobbying profileGeneral web source · 58%

16OpenAI receives the Private Government badge because it combines large-scale private rule-setting over customers, workers, users, members, or counterparties with OpenLobby-documented lobbying of about $680K on public policy issues.

Private Government Verified Medium confidence AI-generated / human-pending May 11, 2026

AI scaffold

OpenLobby — OPENAI lobbying profileGeneral web source · 58%

Competitors

DEXTRACTIVE22 Basecamp private software company · same market DEXTRACTIVE15 Microsoft public company · same market DEXTRACTIVE10 Nvidia public company · same market DEXTRACTIVE8 Oracle public company · same market DEXTRACTIVE7 Adobe public company · same market FMALIGNANT-14 RealPage rental pricing software company · same market FMALIGNANT-22 Palantir Technologies government data and surveillance software company · same market

Help Improve This Profile

Submit source-backed evidence or challenge a specific claim, source, axis value, or calculation below.

How This Page Is Maintained

Evidence comes in through contributors, is checked by verifiers, and is synthesized by reviewers. Founder authority remains narrow and visible; scores recalculate when verified claims or the rubric change.

Civic Note

Incorporation, limited liability, market access, and other institutional privileges are public grants. Good Companies Directory treats those privileges as conditional on accountability to workers, users, communities, and the public.

Submit evidence for OpenAI Add one source-backed fact for review.

Challenge this rating Point to a specific score, claim, source, or calculation problem.

Audit Log

Recent public changes for this company or group. The full audit log is part of the Transparency record.

May 6, 2026 at 12:00 AM UTC ai-scaffolding toxic-products-assessment

Added tentative AI-generated Toxic Products penalty 0 and directory value for current toxic, addictive, pesticide, livestock, chemical, or consumer-safety product exposure.

May 6, 2026 at 12:00 AM UTC ai-scaffolding confidence-calibration

Rechecked evidence coverage and source quality during the evaluation-quality pass; confidence now reflects direct source coverage, complexity, and remaining uncertainty rather than score direction.

May 5, 2026 at 12:00 AM UTC ai-scaffolding ai-scaffolding-score

Added tentative AI-generated claim-backed score components for a previously listed under-review company or group.

View Full Audit Log