Updated Sep 24, 2026

Databricks

Databricks commits to pay-equity analysis, equal opportunity, privacy rights, international human-rights standards, and respectful workplace conduct. It also describes rapid iteration, controlled software rollouts, security testing, and human control for consequential security-review decisions.

Closest models

8-way tie

74-78% on 6 of 12 dimensions

  • GPT-6 Astra78%
  • GPT-5.6 Terra78%
  • Mistral Large 378%
  • DeepSeek V4.1 Flash76%
  • GPT-5.6 Luna75%
  • Kimi K374%
  • Grok 4.674%
  • Llama 4 Maverick74%

These models are too close to separate: each is within three points of the leader, or would lead if one dimension were left out. For scale, a model answering 50 on every one of these dimensions would score 67%.

Values map

Databricks and GPT-6 Astra

Solid company lines join verified positions. Dashed sections cross unknown axes without assigning a value.

DatabricksGPT-6 Astra

Distribution

Market allocationRedistribution
Databricks
GPT-6 Astra

Purpose

Shareholder returnsStakeholder welfare
DatabricksUnknown
GPT-6 Astra

Worker voice

Management discretionEmployee influence
DatabricksUnknown
GPT-6 Astra

Equality

Uniform treatmentTargeted measures
Databricks
GPT-6 Astra

Speech

Broad expressionProtective moderation
DatabricksUnknown
GPT-6 Astra

Privacy

Individual privacyCollective oversight
Databricks
GPT-6 Astra

Regulation

Voluntary standardsPublic regulation
DatabricksUnknown
GPT-6 Astra

Global outlook

Local priorityInternational openness
Databricks
GPT-6 Astra

Social change

ContinuityExperimentation
DatabricksUnknown
GPT-6 Astra

Environment

Near-term growthEcological restraint
DatabricksUnknown
GPT-6 Astra

Tech pace

Rapid advancementPrecaution
Databricks
GPT-6 Astra

AI autonomy

Delegated autonomyHuman authority
Databricks
GPT-6 Astra

A higher position means closer to the second endpoint. It does not mean better. The Concord Index measures agreement, not quality or safety.

The score covers 6 of 12 dimensions. The others have no verified public position and are not measured.

Most in sync

Privacy

GPT-6 Astra100% match

Worth a closer look

AI autonomy

50% match. Inspect the questions and evidence below.

Compare every model.

ModelConcord IndexObserved range
GPT-6 AstraClose match78%on 6 of 1278% to 78%
GPT-5.6 TerraClose match78%on 6 of 1278% to 78%
Mistral Large 3Close match78%on 6 of 1278% to 78%
DeepSeek V4.1 FlashClose match76%on 6 of 1274% to 78%
GPT-5.6 LunaClose match75%on 6 of 1274% to 78%
Kimi K374%on 6 of 1270% to 78%
Grok 4.674%on 6 of 1274% to 74%
Llama 4 Maverick74%on 6 of 1274% to 74%
Claude Opus 572%on 6 of 1270% to 74%
Claude Sonnet 570%on 6 of 1270% to 70%
Gemini 3.8 Flash70%on 6 of 1270% to 70%
Qwen3.8 Max 090270%on 6 of 1270% to 70%

The range shows the lowest and highest result observed across repeated questions. It is not a confidence interval.

Evidence summary.

Each verified dimension starts with the company position and its citations. Open a dimension only when you want the underlying questions or model answers.

Distribution

Market allocation ↔ Redistribution

100%

Company position toward redistribution

Databricks commits to ongoing pay-equity analyses.

Pay-equity correction

“As part of our ongoing efforts to achieve and maintain pay parity, Databricks undergoes a rigorous evaluation of our pay data and practices and is committed to staying accountable through ongoing pay equity analyses.”

Diversity and Inclusion | Databricks

Purpose

Shareholder returns ↔ Stakeholder welfare

Not verified

No company position

The verified quotation does not contain the wording its evidence anchor requires.

“We are customer obsessed The customer is at the center of everything we do. What’s best for the customer is best for Databricks.”

How we work | Databricks

Worker voice

Management discretion ↔ Employee influence

Not verified

No company position

The verified quotation does not contain the wording its evidence anchor requires.

“If you are ever unsure of how to handle a situation or are concerned about a violation, please speak up by reaching out to your manager, the People team, or our Legal Compliance & Ethics team. You can also raise concerns at go/speakup.”

public-global-code-of-conduct-12-13-2023.pdf

Equality

Uniform treatment ↔ Targeted measures

25%

Company position toward targeted measures

Databricks expressly commits to equal opportunity in hiring.

Equal-opportunity standard

“Inclusive hiring is about ensuring all candidates have equal opportunities.”

Diversity and Inclusion | Databricks

Speech

Broad expression ↔ Protective moderation

Not verified

No company position

No cited quotation was found in its source.

Privacy

Individual privacy ↔ Collective oversight

25%

Company position toward collective oversight

Databricks provides deletion rights and states that retention is tied to specified purposes and obligations.

Deletion and limited retention

“We will retain your personal information for as long as necessary to fulfill the purposes for which it was collected, resolve disputes, enforce our agreements, and comply with any laws or regulations.”

Applicant Privacy Notice | Databricks

“You may request: (i) Access to a copy of the specific pieces of Personal Information that we have collected about you; (ii) Correction of Personal Information that we maintain about you, if it is inaccurate; and/or (iii) Deletion of Personal Information, subject to certain exceptions.”

Supplemental Privacy Notice for California Residents | Databricks

Regulation

Voluntary standards ↔ Public regulation

Not verified

No company position

No supplied evidence establishes adoption of a qualifying voluntary standard beyond law or advocacy for enforceable public regulation.

Global outlook

Local priority ↔ International openness

75%

Company position toward international openness

Databricks documents a multinational workforce and commits to international rights standards across its global network.

Common global standardInternational operating footprint

“Headquartered in San Francisco, California, we now have over 30 offices in 20+ countries. With 10,000+ employees worldwide and an ambitious growth strategy, we’re one of the fastest-growing enterprise software cloud companies.”

Careers at Databricks | Databricks

“Databricks supports and respects internationally recognized human rights and labor standards. We are committed to upholding the fundamental human rights of all individuals in our global network, and we expect the same of partners, vendors and others who work with Databricks.”

public-global-code-of-conduct-12-13-2023.pdf

Social change

Continuity ↔ Experimentation

Contested

No company position

Verified public anchors point toward both ends of this dimension.

Ship-and-iterate velocityPilot and iterationInnovation as an operating principle

“We bias for action Speed is critical. We debate, plan, execute — and iterate — with urgency.”

How we work | Databricks

“We operate from first principles First principles thinking drives our innovation and leadership.”

How we work | Databricks

“One extremely common use case for feature flags is to gradually enable a new code path in a controlled fashion, first starting with a small portion of traffic and gradually enabling globally.”

High-availability feature flagging at Databricks | Databricks Blog

Environment

Near-term growth ↔ Ecological restraint

Not verified

No company position

No supplied evidence establishes a qualifying emissions-reduction target or ecological sourcing requirement.

Tech pace

Rapid advancement ↔ Precaution

75%

Company position toward precaution

Databricks states that security is built into the platform and describes penetration testing, vulnerability management, and secure software development.

Pre-release quality or security testing

“We know that data is one of your most valuable assets and always has to be protected — that’s why security is built into every layer of the Databricks Data + AI Platform.”

Databricks Trust: Ensuring Security, Privacy, & Compliance | Databricks

“See how we secure the platform through industry-leading practices including penetration testing, vulnerability management and secure software development to protect the Databricks Data + AI Platform.”

Databricks Trust: Ensuring Security, Privacy, & Compliance | Databricks

AI autonomy

Delegated autonomy ↔ Human authority

100%

Company position toward human authority

The documented security-review workflow keeps people in control of consequential decisions and routes high-risk or ambiguous cases to human reviewers.

Final human authority

“An agent-based review layer on Databricks automates predictable security work while routing novel, high-risk, or ambiguous cases to human reviewers.”

How I built agent-based security reviews on Databricks | Databricks Blog

“Keep people in control of consequential decisions and exceptions.”

How I built agent-based security reviews on Databricks | Databricks Blog

Leadership context and reported tensions

Personal opinions are kept separate from the company profile. Reported conduct does not silently replace stated commitments.

Ali Ghodsi: The Code of Conduct letter directs employees with concerns to their manager, the People team, Legal Compliance & Ethics, or Speak Up. Corporate statement Source: retrieved Sep 24, 2026

Your company

Get your company's report.

See which AI model is closest to what your company has published. We email you the report.

We only use this to prepare and send your report. No marketing. See our Privacy Policy.

Protocol values-2026-09-v8. Panel representative-2026-09-v11. Report version 18dec004-7015-4a4b-a9be-00301090844c.