Is there training done on my data? Does AI learn from data?
No, BlockSurvey’s AI systems do not train on your data, and the AI does not learn from or retain any user inputs. All AI-powered features are designed to operate in a stateless and non-learning mode. Your data is processed in real-time to generate responses but is not stored, logged, or reused for training or model improvement. This ensures your data remains private and is not incorporated into any AI training pipeline, whether internally or by third-party providers.
How is personal information handled with AI? Is it being sent to AI during data analysis?
We implement automated filters to detect and remove Personally Identifiable Information (PII) or Protected Health Information (PHI) before any data is sent for AI processing.In cases where AI is used for data analysis, we utilize a custom machine learning model hosted securely on our own servers. This ensures that data is processed in a controlled, private environment, without being transmitted to external third-party AI providers. Additionally, we do not store or log any data processed through AI, and all processing is done in-memory for added security.
What's the accuracy of AI in data analysis? Are there any bias?
AI accuracy is task-dependent and varies by use case. We use AI models that are carefully tested and optimized for structured data analysis and insight generation. AI models that are carefully tested and optimized for structured data analysis and insight generation. These models perform well in identifying patterns, summarizing responses, and generating actionable insights.
However, like any system built on historical data, AI models may carry inherent biases. We actively monitor outputs and choose providers and models that implement bias mitigation strategies to minimize risks and promote fairness.
How is AI being audited?
We conduct regular internal audits of AI-related features to ensure responsible and compliant use. These audits include reviewing output quality, validating security protocols, and ensuring all usage aligns with privacy regulations such as GDPR, SOC2, and ISO 27001. We are committed to ethical AI practices and transparency in all data processing activities.
What's your AI privacy policy?
You can find our full AI privacy policy here :
https://blocksurvey.io/ai-policy Our AI privacy practices are guided by the following principles:
- We do not use your data to train AI models.
- No personally identifiable information (PII) or protected health information (PHI) is processed as input to AI.
- Data used in AI features is filtered, anonymized where applicable, and processed in-memory.
- We use custom models hosted on our own servers to ensure no third-party data exposure.
- No logging or data retention occurs during AI processing.
Which AI third parties do you use? Do you send data outside?
For AI-based features, particularly data analysis—we use custom machine learning models hosted entirely on our servers. This means no data is sent to external AI providers for processing.
In limited cases, such as AI-assisted question generation, we may integrate with third-party providers like OpenAI. However, in these instances:
- No Personally Identifiable Information (PII) or Protected Health Information (PHI) is shared.
- We apply automated data filtering and minimization techniques before any data is transmitted.
- All communication is done over secure, encrypted channels.
Is there anonymization of data being done in AI before processing?
Yes, BlockSurvey performs anonymization and data minimization before any data is processed by AI.
We have automated filtering mechanisms in place to identify and remove any personally identifiable information (PII) or protected health information (PHI) from the input data. This ensures that only non-sensitive, anonymized content is used during AI processing. Additionally, our AI models are hosted on self-managed servers, and data is processed in-memory without logging or storage, adding an extra layer of protection. These safeguards are designed to uphold our commitment to privacy, confidentiality, and regulatory compliance.
What happens to my survey data with AI analysis?
We take data security and privacy seriously. Your data is encrypted and stored securely. Our AI model processes only the question title and not the data to generate insights, and then it's anonymized and stored securely. It is never shared with third parties.
How is my survey data used?
Your data is only used to generate insights for your surveys. We don't use it for training models or any other purpose.
Do you share my data with partners or affiliates?
No, we don't share your data with anyone. It remains confidential and accessible only to you.
How is your AI Data Analysis different from others?
Our AI Data Analysis is specifically fine-tuned for survey data analysis, using a survey knowledge base to provide more accurate insights. Other generic large language models are general purpose and may not be as precise for survey-related analysis. BlockSurvey is building small language models focused on the survey domain.
Can your AI survey analysis be influenced by the data it's trained on?
Yes, like any AI model, our analysis can be influenced by the data it's trained on. However, we take steps to mitigate this by using diverse, high-quality datasets and regularly updating our models to reflect new information. As we build further, we also have processes to detect and correct any potential biases.
How accurate is your AI-powered survey analysis?
Our AI-powered analysis is highly accurate, with an accuracy rate of 80%. We achieve this through advanced reinforced learning on large datasets, and expert validation. However, like any AI model, ours is not perfect, and we continually work to improve its accuracy.