Company
A swiss-based company 🇨🇭
Founded in Zürich, Switzerland.
Artificial Intelligence Suisse SA, PO 280, Delémont, Switzerland.
Founded in Zürich, Switzerland.
Artificial Intelligence Suisse SA, PO 280, Delémont, Switzerland.
Monitordata privacy
with AI
Use our powerful ai to manage, protect and monitor your data privacy in complete safety.
Cofounders from
Try yourself-Demo showcase
Identification labeling
Ms. North LAST NAME, 88% confidence, our office in Essex confirmed your visit with health psychologist at 10:02 AM.
Labeling for identification involves a detailed process of detecting and categorizing sensitive data entities within your text, ensuring clarity and organization.
At ai4privacy we believe that every company should have a simple and affordable way to protect their data privacy.
What we do-Data privacy made simple
Our AI comes built with the following features in every product we offer.
- Identification. Classify sensitive information with 54 entities across 229 topics in 4 languages.
- De-identification. Extract sensitive information as metadata and anonymize it to secure privacy.
- Re-identification. Use the metadata to re-identify the sensitive information partially or in full when needed.
- Customisation. Adapt every aspect of the identification to your specific needs and requirements.
- Access control. Manage access to the sensitive information metadata with fine-grained permissions lists.
How we do it-Linguistics and Machine Learning at the core
Our core technology is divided into two key parts for efficient recognition of the privacy aspect of data.
Linguistic coverage
An approach to data privacy based on linguistics.
Our linguistics-driven approach identifies 54 key entities from sensitive data to PII, ensuring nuanced contextual understanding.
Covering 229 topics in 5 interaction styles, our model excels in diverse and complex linguistic scenarios.
Multilingual capabilities in English, French, German, and Italian, guaranteeing comprehensive language coverage.
Machine Learning
A model trained on 13.6M+ tokens and 200k+ examples.
We're assessing our model's accuracy, precision, and recall to ensure it reliably identifies and captures sensitive data while minimizing false positives.
These evaluations are key to delivering robust data protection.
What we offer-Product selection
Our products are designed with seamless integration and scalability, enabling you to seamlessly incorporate privacy into your existing workflows and applications.
guardweb extension
Browse safely
Browse the web safely with our browser extension that automatically detects and anonymizes sensitive data.
filterintegration
Anonymize your data
Mask the Personal Identifiable Information of your internal data to protect your customers privacy with anonymization.
datasetlicense
Use our dataset commercially
Get access to our dataset of 200.000+ examples to train your own models.
your requestcustom
Request custom solution
We can help you with your specific needs and infrastructure. Contact us to discuss your requirements.
Schedule a call with us
Who we are-About us
We are a team with a passion for data privacy and machine learning coming from the academic and industrial world.
Cofounders from ETHZ, Google and CERN.
Open‑Source Community
We are committed to open-source development.
Our dataset is proudly open-source, reflecting our commitment to raising awareness about privacy and giving back to the community with tools that are free for personal use.
You can find us on Hugging Face.