Grok
xAI · United States
What it is
xAI’s AI assistant, positioned around real-time data from X (formerly Twitter). It drafts and answers, runs a DeepSearch web/X search, and offers voice and image and video generation (Grok Imagine). It comes three ways: built into X, as standalone apps at grok.com and on mobile, and through an API.
How it works
Grok is built on xAI’s Grok family of large language models (the point versions move quickly, so we describe the family). Its distinctive feature is tight integration with live X data. The three access paths matter for your data: Grok inside X is governed by X’s terms; the grok.com app and the API are governed by xAI’s.
Who is behind it
- Owner
- SpaceX. Originally founded as an independent entity in March 2023, xAI acquired its sister company X Corp. in March 2025. In February 2026, SpaceX acquired the combined operation. xAI was subsequently dissolved as a standalone corporate entity and consolidated into SpaceX as its dedicated SpaceXAI division.
- Founded
- xAI was founded on 9 March 2023 (announced July 2023).
- Headquarters
- San Francisco Bay Area, California, United States — xAI now operates as the SpaceXAI division of SpaceX. Consumer and grok.com/API terms are governed by Texas law; enterprise agreements by Tennessee law.
Plans
- Grok on X (Free / X Premium / Premium+) and grok.com / Grok apps (Free / SuperGrok) · Consumer (free / paid)Grok on X trains on your prompts and public X posts by default; the standalone apps offer an opt-out and Private Chat, but the NZ default is contested.
- xAI Enterprise · Enterprise / WorkspaceNo training without permission; ZDR, SOC 2 and BAA options. Governed by Tennessee law.
- xAI API · API directNo training without permission; ~30-day default retention.
How it handles your data
Consumer Grok is the weakest of the mainstream consumer tools for confidential use. Used inside X, it trains on your prompts and on your public X activity by default, and X’s licence over your content is broad (worldwide, and able to be sub-licensed, including for AI training); the opt-out is a buried setting and is not retroactive. On the standalone grok.com app and Grok apps the position is better in principle — there is an opt-out and a Private Chat mode — but the training default for users outside the EU (including New Zealand) is genuinely contested in the current documents, so we treat it as training-by-default until confirmed. By contrast, the xAI API and Enterprise agreements do not train on your content without permission, default to roughly 30-day retention, and offer Zero Data Retention (Enterprise), with SOC 2 and BAA options. There is no New Zealand or Australia data region.
Sources for this description
- X Privacy Policy (governs Grok on X) — X Corp.live sourcePerma.cc recordas at 31 May 2026
- About Grok — data use & training opt-out — X / xAIlive sourcePerma.cc recordas at 31 May 2026
- xAI Terms of Service (consumer) — xAIlive sourcePerma.cc recordas at 31 May 2026
- xAI Consumer FAQs (training / retention / settings) — xAIlive sourcePerma.cc recordas at 31 May 2026
- xAI Terms of Service (enterprise) — xAIlive sourcePerma.cc recordas at 31 May 2026
- xAI Privacy Policy — xAIlive sourcePerma.cc recordas at 31 May 2026
- About xAI (company facts) — xAIlive sourceas at 31 May 2026
What is safe, on which plan
Each cell summarises what the provider’s terms and the applicable rules say. The colour is our summary of the sources, not advice — select a cell to read the reasoning and follow the sources yourself.
- Green
- Amber
- Red
- Black
- Not yet assessed
| Data category | Consumer (free / paid) | Enterprise / Workspace | API direct |
|---|---|---|---|
| 1. Public information | |||
| 2. Internal non-confidential | |||
| 3. Internal confidential (non-privileged) | |||
| 4. Legally privileged | |||
| 5. Personal information (Privacy Act 2020) | |||
| 6. Sensitive personal information | |||
| 7. Children’s information | |||
| 8. Court-protected material |
1. Public information
- Consumer (free / paid)
- Enterprise / Workspace
- API direct
2. Internal non-confidential
- Consumer (free / paid)
- Enterprise / Workspace
- API direct
3. Internal confidential (non-privileged)
- Consumer (free / paid)
- Enterprise / Workspace
- API direct
4. Legally privileged
- Consumer (free / paid)
- Enterprise / Workspace
- API direct
5. Personal information (Privacy Act 2020)
- Consumer (free / paid)
- Enterprise / Workspace
- API direct
6. Sensitive personal information
- Consumer (free / paid)
- Enterprise / Workspace
- API direct
7. Children’s information
- Consumer (free / paid)
- Enterprise / Workspace
- API direct
8. Court-protected material
- Consumer (free / paid)
- Enterprise / Workspace
- API direct
Select any cell to see what the sources say and why.
Statements about provider terms, plans, and defaults will go stale. This profile names its last review date above.
Where to next
Our framework
The reference behind every answer: the eight kinds of information, the five plans, and what each colour means.
Our frameworkGuidance
Our view on the four questions to ask before you use a tool — confidentiality, privilege, privacy, and the professional rules.
GuidanceAbout us
Who is behind this, how we keep it reliable, and why you can trust it.
About us