What is the EU AI Act?
The EU AI Act (Regulation (EU) 2024/1689) is the European Union's risk-based regulatory framework for artificial intelligence. It classifies AI systems into four tiers:
Unacceptable risk - practices banned outright, including certain uses of emotion recognition in the workplace (Article 5)
High-risk - systems subject to strict obligations around human oversight, documentation, and risk management (Annex III)
Limited risk - systems subject to transparency obligations, such as disclosing that a user is interacting with AI or that content is AI-generated (Article 50)
Minimal risk - no specific obligations
The Act also assigns different obligations depending on role: the provider develops an AI system and places it on the market; the deployer uses that AI system under its own authority in a professional context.
Bliro is the provider of the AI features described in this document. When a customer uses Bliro's AI features, their organization is the deployer, and for most configurations, deployer obligations are minimal.
How Bliro Maintains Compliance
Bliro's core product is AI-powered transcription, summarization and personal assistants. As the provider, Bliro offers the following capabilities to maintain compliance.
Transparent Labels on all AI-Generated Content*
Low-Risk Defaults for all Features
In-App Warnings on Higher-Risk Configuration Choices
No Biometric-Based Emotion Recognition
Human Oversight Controls for Agent Actions
* Per EU AI Act, Transparent Labels for Machine-Readable content is due December 2, 2026. While every feature may not exist at the time of writing, Bliro commits to meet both of these timelines.
How to Leverage Bliro as a Low-Risk Platform
As the deployer, the customer is in control of how these features are leveraged by their organization. Below is a list of actions that may generally classify as Low-Risk. Please note that this document is not a substitute for internal review of customer internal processes. Customers should consult their internal compliance teams as needed to ensure compliance.
Meeting Transcription and Summarization
User-Guided Creation of Meeting Follow-ups, including Emails, Meeting Invites and CRM Updates.
Interactive Agent Sessions to Analyze Meeting Transcript Content
VoiceID for Automatic Detection of the User’s Voice in Personal Mode
Coaching Analysis for Sentiment Detection based on Transcript Content
Optional, Potentially High-Risk Features
Some Bliro capabilities can be configured beyond their low-risk default. Enabling these may shift you into the role of deployer of a high-risk AI system under the Act, with the corresponding obligations (human oversight, informing affected individuals, and related documentation).
VoiceID in Organization or Global Mode
Organization-scoped or Global-scoped recognition. Beyond the 1:1 default, an organization can opt in to broader voice recognition (an enrolled user identified across other users' calls, org-wide, or globally). Under EU AI Act, this may fall into biometric identification under Annex III point 1. In this case, the feature would classify high-risk. To inform organizations about the classification, Bliro actively informs the admin user when this broader scope is selected.
Coaching and performance analytics
Individual Employee Evaluation. Where these features are configured or used by an employer to evaluate the performance of identifiable individual employees - for example, performance tracked per named sales representative - that use can fall within Annex III point 4 (AI systems for monitoring or evaluating worker performance). In that scenario, you act as the deployer and carry the corresponding deployer obligations.