“Is Perplexity GDPR-compliant?” is the question a data protection officer gets after the research team has been pasting customer briefs into Perplexity Pro for a month. The honest answer is: it depends on the plan, the contract and where the data is processed, not on the brand. This article walks through what Perplexity offers today, what remains open for a company in Germany, Austria or Switzerland, what it costs and how to start - factually, as of October 2026. This is not legal advice; confirm the details with your data protection officer.
Key takeaways
The plan decides: Perplexity Enterprise runs with a Data Processing Addendum and the commitment that content is not used to train models. Free, Pro and Max are consumer accounts where AI data retention is on by default and the opt-out only covers future data123.
US hosting, no EU option on offer: Perplexity is a US company; its notice says data may be processed in the US. A dedicated EU inference or storage region is not published1.
Your duties stay with you: lawful basis (Art. 6 and 9 GDPR), record of processing (Art. 30), DPIA (Art. 35) where the risk is high, plus AI literacy under Art. 4 of the EU AI Act89.
Is Perplexity GDPR-compliant?
Perplexity is not GDPR-compliant or non-compliant on its own: on an Enterprise plan with the DPA, clear rules for your team and your own documentation, companies can use it in a GDPR-compliant way. Perplexity Enterprise says it never trains models on customer content and offers SOC 2 Type II, SSO and SCIM, audit logs and configurable file retention. The open points are the private accounts in the team, the US processing location and the paperwork that stays with you123.
Two worlds under one name
Perplexity separates consumer plans (Free, Pro, Max), governed by its Privacy Notice without a DPA, from Enterprise and the API, governed by the Enterprise or API Terms with the Data Processing Addendum built in134. Same search engine, different contract.
The EU-first path: Superkind
If you want a research-grade answer that then becomes real work, from a tool hosted in Europe, Superkind is one way. Superkind is the AI employee for your team: you write to it in Outlook or any email, Microsoft Teams, Slack or the Superkind web app like you would to a colleague, and it does the task in your programs through 1,500+ integrations such as SharePoint, DATEV, HubSpot and Personio. Superkind is hosted in Europe, GDPR compliant, aligned with the EU AI Act and never trains models on your data. It uses AI models from OpenAI, Anthropic, Mistral and Google, processed in Europe, and the DPA comes through the Trust Center.
- 1
You write the task
In Outlook, Teams, Slack or the web app, in plain words: “Research our top three competitors in Austria, pull the latest from their websites and press, and summarise the pricing changes.”
- 2
Superkind does it in your systems
It searches the open web, pulls from your SharePoint and HubSpot, follows your rules such as “Competitor notes go into the HubSpot account record” and uses your Company Brain so it knows which three competitors matter.
- 3
Superkind asks before anything goes out
You get the summary with sources and the draft HubSpot note with Approve or Reject. Nothing is saved to the CRM or sent to colleagues without your OK, and a teammate can approve in Teams or Slack.
- 4
Next time it runs as a routine
What you explain once keeps running, for example “Every Monday, check the three competitors and post the diff in Teams”. Every task adds to your Company Brain.
Research the three Austrian competitors from our HubSpot list, pull the latest pricing from their websites and the trade press, and draft a one-page summary. Send it to me before anything goes in HubSpot.
Send Superkind your first research task
Start for free in Outlook, Teams or Slack. Hosted in Europe, no training on your data.

What Perplexity offers today
Six facts from Perplexity’s own pages decide how it fits the GDPR.
- A consumer product and an Enterprise product - the Privacy Notice governs Free, Pro and Max; the Enterprise and API Terms of Service govern Enterprise and the developer Sonar API. The DPA is incorporated by reference into both of those Terms134.
- No training on Enterprise content - the Enterprise Terms (Section 2.3.3) commit Perplexity not to use Customer Content to train, retrain, fine-tune or otherwise improve generative AI models, and not to authorise third parties to do so. The product page says the same: “We never train our LLMs on enterprise customer’s data”23.
- Training on by default for consumer plans - on Free, Pro and Max, the AI data retention setting is on by default. The opt-out is in Preferences and only covers future data; logged-out searches are also used by default7.
- Admin and security on Enterprise - the product page lists SOC 2 Type II certification, SSO plus SCIM, audit logs for compliance and configurable file retention with automatic deletion down to one day. It also states Perplexity is “GDPR and HIPAA compliant”2.
- US hosting, no EU region on offer - the Privacy Notice says data may be processed in other countries, “including, but not limited to, the United States.” An EU inference or storage region is not published on Perplexity’s pages as of October 20261.
- Models under the hood - Perplexity routes queries through its own Sonar model, built on Llama 3.3 70B, plus frontier models from OpenAI (GPT-5-class) and Anthropic (Claude 4.5-class). On Enterprise, none of these see your content for training36.
What is missing for company use
The contract is in place on Enterprise. The gaps are in practice and in the paperwork.
- Private Pro accounts in the team - many people started with Free or Pro. Those searches sit outside your DPA, and the AI data retention toggle is on by default; the opt-out only covers future data7. Without a rule, customer briefs and prospect names end up in that pool.
- US processing without an EU option - if your data protection policy requires processing in the EU, Perplexity does not offer that choice on its own pages; data may be processed in the United States1. Transfers to third countries have to be covered by Standard Contractual Clauses and named in the record of processing8.
- A works council question - audit logs, admin consoles and SSO on Enterprise are good for IT, and in Germany they are also an item for the Betriebsrat under Section 87 (1) 6 BetrVG. Agree the logging and monitoring scope before rollout.
- Answers, not actions in your programs - Perplexity gives you a researched answer with citations. It does not book an invoice in DATEV, create a HubSpot deal or send an approved email through Outlook. For work that moves data between systems with an approval step, you need a different tool, or you add one next to it.
- Your own documentation - a lawful basis under Art. 6 (and Art. 9 for health or other special data), an entry in the record of processing (Art. 30) and a DPIA (Art. 35) where the use is likely to be high risk8.
- AI literacy and transparency - Art. 4 EU AI Act has required AI literacy for staff since 2 February 2025; Art. 50 adds transparency duties from 2 August 2026, such as labelling AI-generated text published to inform the public unless a person has reviewed it9.
Perplexity plans and the GDPR compared
As of October 2026. Where each option stands on the points that decide GDPR fit, with Superkind for comparison, based on the vendors’ own pages1237. Not a compliance sign-off.
| Option | Works with | Training on your data | Hosting / Data protection | DPA |
|---|---|---|---|---|
| Superkind (for comparison) | Outlook, Teams, Slack, web app; does tasks in 1,500+ integrations (SharePoint, DATEV, HubSpot ...) with approval | Never | Hosted and processed in Europe; models from OpenAI, Anthropic, Mistral and Google | Yes, via the Trust Center |
| Perplexity Free, Pro, Max | Chat and research; one person; limited files and uploads | On by default; opt-out in Preferences7 | Consumer Privacy Notice; US processing possible, no EU region listed1 | No |
| Perplexity Enterprise Pro | Chat, Spaces, files, admin console; SSO and SCIM, audit logs | Never (contractual)23 | Enterprise Terms; SOC 2 Type II; US processing possible, no EU region listed12 | Yes, in the Enterprise Terms3 |
| Perplexity Enterprise Max | As Enterprise Pro, plus higher usage and newest frontier models | Never (contractual)23 | Enterprise Terms; SOC 2 Type II; US processing possible, no EU region listed12 | Yes, in the Enterprise Terms3 |
| Perplexity Sonar API | Your own software and agents | Never (API Terms)4 | API Terms; US processing possible, no EU region listed1 | Yes, in the API Terms4 |
In short: Free, Pro and Max are the wrong place for customer or employee data; with Enterprise or the Sonar API you get a DPA and no training, and the open point is the US processing location. For the same question on other tools, see Is ChatGPT GDPR-compliant?, Is Claude GDPR-compliant? and Is Microsoft Copilot GDPR-compliant? For a shortlist of alternatives in the same niche, see Perplexity alternatives.
Examples: Perplexity and the GDPR in practice
Five situations and what each one needs to stay compliant.
1. Superkind runs the research task, in Europe
Nadine asks Superkind in Teams to research the three Austrian competitors and prepare a HubSpot note. Superkind pulls from the public web and your 1,500+ integrations such as SharePoint, HubSpot and DATEV, follows the rules in your Company Brain and asks her before anything goes into the CRM. Processing happens in Europe, there is no training on the data, and the DPA comes through the Trust Center. See AI sales rep for how this runs day to day.
2. The strategy team on Perplexity Enterprise
Five people in strategy use Perplexity Enterprise. The DPA applies, your content is not used for training, SSO controls who gets in and audit logs record use23. What is left: add Perplexity to your record of processing, name the transfer to the US, agree the logging scope with the works council, and write a rule for which data goes in.
3. A colleague uses Perplexity Pro privately
A sales manager pastes a customer brief into a personal Pro account. That search is outside your DPA, and training is on by default unless the person switched it off; even then, data already used for training is not removed7. Make the rule explicit: company data only in company accounts.
4. Research with health data
An HR lead wants Perplexity to summarise return-to-work plans. Health data is a special category under Art. 9 GDPR, so a DPIA under Art. 35 is likely due before use8. For how to run one, see DPIA for AI agents.
5. Developers build on the Sonar API
The IT team builds an internal research assistant on the Sonar API. The API Terms incorporate the DPA, usage data is permitted for Perplexity but not Customer Content, and the processing location is still the US14. Add Perplexity to the record of processing and plan for the SCCs.
What a compliant Perplexity setup needs (and what it costs)
Five points, none of which happens by itself.
- Enterprise accounts only - company data only in Enterprise or the Sonar API, where the DPA applies and training is contractually off34.
- A rule for private accounts - no customer or employee data in Free, Pro or Max; move active users to Enterprise or to a tool that processes in Europe.
- A decision on the data location - accept US processing with Standard Contractual Clauses and name the transfer in your record of processing, or move to a tool that processes in Europe18.
- Record of processing and DPIA - add Perplexity to your record (Art. 30) and run a DPIA (Art. 35) where the risk is high8.
- AI literacy, transparency and works council - train staff and set rules under Art. 4 EU AI Act; agree logging and monitoring with the Betriebsrat before rollout; label AI-generated public text under Art. 509.
What Perplexity costs
As of October 2026. List prices in US dollars, excl. tax, from Perplexity’s pricing page and third-party reviews2. Enterprise is sold as Enterprise Pro and Enterprise Max; a full custom contract exists for larger buyers.
| Plan | Price | Note |
|---|---|---|
| Superkind (for comparison) | Free start, Business from EUR 50 per month | For the whole company, unlimited users, no price per seat |
| Perplexity Free | $0 | Limited Pro searches; consumer plan, no DPA |
| Perplexity Pro | $20 per month, $200 per year | Personal plan, no DPA; training on by default7 |
| Perplexity Max | ~$200 per month | Heavy use, newest frontier models, personal plan |
| Perplexity Enterprise Pro | ~$40 per seat per month (list) | SSO, SCIM, audit logs, no training on your data2 |
| Perplexity Enterprise Max | ~$325 per seat per month (list) | Enterprise Pro plus highest usage and top models |
Ten seats on Enterprise Pro at list come to around $400 per month before tax, plus Sonar API usage if you build on top. Perplexity publishes its own custom Enterprise pricing on request; third-party sources agree on the $40 figure but disagree on annual discounts, so confirm with sales.
Data protection and EU hosting
Safe by default
Your data stays in Europe, under a DPA, with no training on your data.
- Hosted and processed in Europe
- GDPR compliant
- Aligned with the EU AI Act
- No training on your data
- DPA through the Trust Center
For a prospect name, a customer brief or an employee file, the personal data is processed in Europe from the first task, even when the heavy lifting runs on an OpenAI or Anthropic model, with approval before anything leaves the team.
What Superkind costs
As of October 2026. Prices excl. VAT, for the whole company with unlimited users.
- Free - EUR 0 with a one-off up to EUR 150 in credits to start. No credit card, no sales call, every feature and connection.
- Business - from EUR 50 per month for 25,000 credits. EUR 100 for 50,000 credits is the popular step, then EUR 200 and EUR 500. No price per seat, unused credits carry over one month, custom apps included, cancel monthly.
- Enterprise - custom price for your volume, dedicated support with SLA, SSO/SCIM and audit log, a DPA and help with the security review, payment by invoice.
- What credits buy - a question uses 10-50 credits, a task 50-500 credits and a run of your own agent 1,000-5,000 credits. EUR 100 covers up to 1,000 tasks or 5,000 questions.
How to start
Start with one real task and put Perplexity in order in the same week. Three steps:
Your 3-step plan
- Test an EU-first AI employee on one real task - start Superkind for free at app.superkind.ai with up to EUR 150 in credits, send one real research task in Outlook, Teams or Slack, and compare the output to Perplexity, processed in Europe. For more options in the same niche, see Perplexity alternatives.
- Write the rule - company data only in Enterprise accounts; move private Pro users to Enterprise or to Superkind; switch off AI data retention for every consumer account that stays; add Perplexity to your record of processing. That also starts your AI literacy duty under Art. 4 EU AI Act9.
- Decide on the data location - if processing in the EU is a must, use a tool that processes in Europe. Compare the options in GDPR-compliant AI tools with EU hosting.
From the first task to your own app
With Superkind, the data stays in Europe from the first task to the app you build later. Four stages, on the research example from above:
- 1
Ready in minutes
Sign up for free, send the first research task in Teams or by email and watch Superkind pull a sourced summary for the three competitors.
- 2
Tasks become routines
What you explain once keeps running as a routine, for example every Monday’s diff for the three competitors in Teams. Before anything goes out, Superkind asks for approval.
- 3
Learns from day one
Every task makes Superkind smarter. Your competitors, win themes and the one-pager format end up in your Company Brain, and that knowledge is never lost.
- 4
Your own app, without starting over
When it gets more complex, you build custom solutions, apps and integrations on top of what Superkind already knows, for example your own competitor radar app with win/loss analysis and HubSpot sync: your Super App. Built yourself or together with us. It all belongs to you.
Conclusion
Perplexity can be used in a GDPR-compliant way: Enterprise comes with a DPA, no training on your content, SOC 2 Type II and SSO. The open points are the private Pro accounts, the US processing location and the paperwork that stays with you. If you want research-grade answers that also do real tasks in your systems, processed in Europe with approval before anything goes out, Superkind connects to SharePoint, DATEV, HubSpot and 1,500+ integrations from EUR 50 per month for unlimited users (as of October 2026). Write your Perplexity rule this week, then send Superkind one real task and compare.
Frequently Asked Questions
Perplexity can be used in a GDPR-compliant way on an Enterprise plan, where the Data Processing Addendum applies, no training on your content takes place and admin controls are in place. Free, Pro and Max are consumer accounts: the AI data retention setting is on by default and the opt-out only covers future data. In every case Perplexity is a US company and personal data may be processed in the United States, so Standard Contractual Clauses apply and the record of processing has to name the transfer. As of October 2026, not legal advice.
Yes, for Enterprise. The Enterprise and API Terms of Service incorporate the Data Processing Addendum by reference, and Perplexity commits in its Enterprise Terms not to use Customer Content to train or fine-tune generative AI models. For individual Pro or Max accounts there is no DPA. For GDPR requests, Perplexity names VeraSafe as its Article 27 representative and data protection officer in the EU and UK.
On Enterprise, no: Perplexity states "We never train our LLMs on enterprise customer's data." On Free, Pro and Max, the "AI data retention" setting is on by default, which allows training on your queries and content; each user has to switch it off in Preferences, and that only covers future data. Logged-out searches are also used by default. If you connect the Email Assistant, Perplexity states it does not use email content to train models.
Perplexity is a US company. Its privacy notice says data may be processed in other countries "including, but not limited to, the United States." A specific EU inference or storage region is not named on its public pages as of October 2026. If your data protection policy requires processing in the EU, Perplexity itself does not offer that choice; a European alternative such as Superkind processes in Europe from the first task.
Perplexity routes queries through its own Sonar model (based on Llama) plus frontier models from OpenAI (GPT-5-class) and Anthropic (Claude 4.5-class). Which models are available depends on the plan, and the lineup changes often. On Enterprise, your content is not used to train any of these models.
Superkind is hosted and processed in Europe, is GDPR compliant, aligned with the EU AI Act and never trains models on your data. It uses AI models from OpenAI, Anthropic, Mistral and Google, processed in Europe, and the DPA comes through the Trust Center. Your team writes to Superkind in Outlook, Teams, Slack or the web app, and it does the task in your programs through 1,500+ integrations such as SharePoint, DATEV, HubSpot and Personio, asking for approval before anything goes out.
As of October 2026, Perplexity Pro is $20 per month ($200 per year), Max is around $200 per month, Enterprise Pro is around $40 per seat per month and Enterprise Max around $325 per seat per month at list price. Superkind starts free with a one-off up to EUR 150 in credits; Business starts from EUR 50 per month for the whole company with unlimited users, no price per seat.
Sources
- Perplexity Privacy Notice (last updated October 2026)
- Perplexity Enterprise - product page (SOC 2 Type II, no training on enterprise data, SSO + SCIM, audit logs)
- Perplexity Enterprise Terms of Service - DPA, no training, usage data
- Perplexity API Terms of Service - DPA incorporated by reference
- Perplexity Trust Center
- Perplexity Blog - Meet the new Sonar (Sonar built on Llama 3.3 70B)
- Perplexity Help Center - Data collection at Perplexity
- EUR-Lex - Regulation (EU) 2016/679 (GDPR), Art. 6, 9, 28, 30, 35
- EUR-Lex - Regulation (EU) 2024/1689 (EU AI Act), Art. 4 and Art. 50
- Superkind Trust Center - DPA, EU hosting, no training
Send Superkind your first task.
Start for free in Outlook, Teams or Slack, hosted in Europe. Or see in 30 minutes how it runs in your systems.

