NetQnect AI Lawful Basis Summary

Version date: 16 September 2026.

Effective date: 16 September 2026, applying only when this exact revision is published.

Contact: legal@netqnect.com

1. About This Summary

This summary explains the lawful bases NetQnect uses for AI-assisted features. It should be read with the Privacy Notice and AI Processing Notice.

2. AI Purposes And Lawful Bases

User-requested AI Assistant chat

• Purpose: answer user requests, draft follow-ups, explain relationship context and help users use NetQnect.
• Lawful basis: contract, where the user asks NetQnect to provide the assistant feature.
• Supporting basis: legitimate interests for safety, debugging, abuse prevention and service reliability.
• User control: users choose what to send to the assistant and can request deletion or export through privacy rights channels.

User-requested profile, bio and content assistance

• Purpose: help users create, edit or improve profile content.
• Lawful basis: contract, where the user requests the generated or edited content.
• User control: users choose whether to use generated content and can edit or delete it.

Embeddings for discovery, search and retrieval

• Purpose: convert profile, event or relationship text into mathematical representations for relevant search, matching and retrieval.
• Lawful basis: legitimate interests, subject to transparency, objection handling, minimisation and deletion controls.
• User control: users can object to non-essential profiling and request deletion through privacy rights channels.

AI matching, recommendation conversion measurement and relationship insights

• Purpose: suggest relevant people, events, actions, follow-ups and networking opportunities; record bounded stateful recommendation outcomes; prevent duplicate conversions; and produce relationship or network insights.
• Lawful basis: legitimate interests, subject to transparency, objection handling and human review routes.
• Supporting basis: contract where the user directly requests the recommendation or resulting feature.
• User control: users can object to non-essential profiling, choose whether to act on recommendations, provide feedback, export Firestore source/outbox facts and use account deletion to trigger downstream projection cleanup. Optional Google/Firebase Analytics consent is separate from canonical recommendation-outcome processing.

Event attendee recommendations and event insights

• Purpose: help attendees and organisers find relevant connections and understand event activity.
• Lawful basis: legitimate interests for recommendations and event intelligence, with contract where a user directly requests an event feature.
• User control: users can object to non-essential profiling and organisers must have rights to upload or connect attendee data.

AI traces, safety review and debugging

• Purpose: detect abuse, debug AI failures, review tool calls and maintain safe service operation.
• Lawful basis: legitimate interests.
• User control: traces are minimised, redacted where practical, and retained for short periods as set out in the Data Retention Schedule.

Legal, regulatory and rights-request handling involving AI records

• Purpose: comply with legal obligations, respond to rights requests, protect legal rights and handle complaints.
• Lawful basis: legal obligation or legitimate interests, depending on the case.

3. Legitimate Interests Safeguards

Where NetQnect relies on legitimate interests for AI processing, it applies:

• clear public transparency in the Privacy Notice and AI Processing Notice;
• a documented Legitimate Interests Assessment;
• minimisation of data sent to AI and embedding providers;
• retention limits for traces and logs;
• redaction of secrets and sensitive contact details where practical;
• objection handling for non-essential profiling;
• deletion and export coverage for AI records;
• human review where a user challenges a recommendation or AI output.

4. Special-category Data

NetQnect is not designed to collect special-category data for AI features. If a user submits special-category data or AI processing creates sensitive inferences, NetQnect must establish the additional condition required by applicable law before that processing. A user entering the data does not itself supply that condition or remove our responsibilities.

5. Automated Decision-making

NetQnect AI features are intended to provide recommendations and assistance. They are not intended to make solely automated decisions with legal or similarly significant effects, such as decisions about employment, credit, housing, insurance, healthcare, pricing or access to essential services.

If any AI output later affects eligibility, access, pricing, ranking with material impact or opportunity allocation, we assess the change and provide the additional user rights controls required by law before introducing it.

6. Purpose Changes And Pseudonymisation

The basis for providing an assistant response does not automatically cover a different later purpose or indefinite retention. A new use needs assessment before it begins, including the original purpose and expectations, data sensitivity, likely effects and safeguards. Pseudonymisation reduces risk but does not remove data-protection duties. If a use is based on consent, withdrawal cannot be bypassed by simply relabelling the same use as legitimate interests.

7. Governance And Optional Programmes

The business owner approved the core Data Protection Impact Assessment on 16 September 2026. It covers third-party information supplied by users and inference-time provider processing, with notices, acceptable-use conditions, provider terms and a non-user notice in organiser attendee import. We are moving inference from Nebius Token Factory to Google Gemini through the EU Vertex endpoint, with the Google Cloud Data Processing Addendum and applicable international-transfer safeguards. The AI Processing Notice and Subprocessors list explain the providers used during this transition. We keep our assessments and rights controls under review.

We do not use your data to train AI models. If we ever propose an optional programme, we will explain it in Privacy Settings and ask you first.