Fireflies.AI Meeting Assistant Allegedly Collected Biometric Voice Data from Illinois Meeting Participant Without Consent

On November 18, 2025, Fireflies.AI reportedly recorded and analyzed the voice of Illinois resident Katelin Cruz during a virtual meeting hosted by Western Illinois Dreamers, although Cruz had no Fireflies account and had not provided written consent. Her lawsuit alleges the system used speaker-recognition functions to create and retain biometric voice data and attribute statements to her in the meeting transcript.

Risk classification

• Primary risk domain: 2 Privacy & Security
• Primary risk subdomain: 2.1 Compromise of privacy by obtaining, leaking or correctly inferring sensitive information

The incident involves the unauthorized collection, analysis, and storage of sensitive biometric voiceprints from non-consenting meeting participants.

Causal factors

• Entity: AI
• Intent: Intentional
• Timing: Post-deployment

The privacy risk is caused by the AI system's automated action of recording and analyzing voices to create biometric voiceprints post-deployment, which was an intentional design feature of the tool.

EU AI Act risk tier

• Risk tier: 3 Limited Risk

Limited Risk: The AI meeting assistant requires specific transparency obligations, such as notifying users when they are interacting with an AI system, which is the central compliance failure alleged in the lawsuits.

AI system and alleged parties

• AI system: Fireflies.AI meeting assistant (Fireflies.AI Corp.)
• AI purpose: Voice Recognition; Writing Assistant
• Behaviour type: Assistant
• Alleged developer: Fireflies.AI, AI transcription technology developers
• Alleged deployer: Western Illinois Dreamers, Organizations, AI transcription tool users
• Alleged harmed parties: Privacy, People recorded without consent, Katelin Cruz, Biometric data subjects

Harm severity

Severity ratings are based on this scale.

Highest direct severity in any category: Minor. Severity is scored from Negligible to Catastrophic in each harm category, for harm the reports describe as caused directly or indirectly by the AI system.

• Physical: direct Negligible, indirect Negligible
• Infrastructure: direct Negligible, indirect Negligible
• Property: direct Negligible, indirect Negligible
• Financial: direct Negligible, indirect Negligible
• Environmental: direct Negligible, indirect Negligible
• Malicious content: direct Negligible, indirect Negligible
• Differential treatment: direct Negligible, indirect Negligible
• Civil rights: direct Negligible, indirect Negligible
• Democracy: direct Negligible, indirect Negligible
• Privacy: direct Minor, indirect Negligible
• Psychological: direct Negligible, indirect Negligible
• Epistemic: direct Negligible, indirect Negligible
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

Privacy

Reported: The report explicitly describes privacy violations involving the unauthorized collection and storage of biometric voiceprints from meeting participants.

Directly caused: Fireflies.AI directly collected, analyzed, and stored the unique voiceprints of Katelin Cruz and other meeting participants without their written consent or knowledge.

Inferred additional harm: It is highly likely that thousands of other unregistered meeting participants had their biometric voiceprints harvested and stored without consent across numerous virtual meetings facilitated by Fireflies.AI.

People affected

• Occurrences reported: 1
• People reportedly harmed: 1
• People reportedly exposed: 1

Potential causes

Management

• Prioritizing Growth Over Safety: Features like stealth recording are marketed to drive competitive growth.
• Lack of Vendor Due Diligence: Organizations deploy AI tools without auditing vendor data practices.
• Deficient Corporate AI Policies: Companies fail to establish clear rules on AI notetaker deployment.

Technology

• Stealth by Design Architecture: Granola runs without visible bots or notifications to meeting participants.
• Biometric Voiceprint Generation: Fireflies.ai speaker recognition creates voiceprints without user consent.
• Viral Credential Propagation: Otter.ai auto-joins and propagates through calendar and email integration.

Data Inputs

• Default AI Model Training: Granola uses captured communication data for default model training.
• Non-Consenting Participant Audio: Audio inputs are captured from meeting participants who never agreed to terms.
• Irreversible Data Incorporation: Once incorporated into AI training, data cannot be extracted or deleted.

Human Factors

• Lack of Participant Awareness: Non-user participants have no way of knowing an AI tool is recording them.
• Host Compliance Neglect: Meeting hosts activate tools without informing other participants.
• Social Pressure to Consent: Participants feel pressured to allow recording rather than disrupt meetings.

Process and Methods

• Inadequate In-Meeting Notice: Tools lack default active notifications like watermarks or chat alerts.
• Absent Deletion Workflows: Vendors lack clear procedures to delete data after access revocation.
• Shifting Liability to Users: Vendors transfer compliance and consent obligations entirely to customers.

Regulatory Environment

• Strict Biometric Regulations: Strict laws like Illinois BIPA penalize unauthorized voiceprint collection.
• Multi-Jurisdictional Patchwork: Varying consent laws create complex legal compliance challenges for companies.
• Inadequate Privacy Law Structure: Existing laws fail to protect non-purchasers whose data is captured.

Information quality

• Classification confidence: High
• Reason for confidence: The reports provide consistent and detailed information regarding the class-action lawsuit against Fireflies.AI, including the specific legal claims under BIPA, the mechanics of the voiceprint collection, and the details of the plaintiff's experience.
• Ambiguities identified: The exact technical process of how voiceprints are generated and whether they are permanently stored or easily deletable remains a point of legal dispute.
• Alternative interpretations: The defense might argue that the tool is a mere transcription utility rather than a biometric identification system, which would shift the classification from a biometric privacy violation to a standard wiretapping issue.

A class-action lawsuit against Fireflies.AI highlights emerging civil privacy concerns regarding unauthorized biometric voiceprint collection by ambient AI assistants. While presenting notable regulatory and privacy challenges for commercial operations, the incident has negligible direct national security implications.

• Overall national security impact: Minor
• Scope: Single nation
• Primary target: United States
• Alleged perpetrator: Fireflies.AI Corp.

Threat characteristics

• Imminence: Long-term. This is a developing legal and regulatory compliance matter, not an active national security threat or immediate crisis.
• Autonomy: Human-supervised. The AI assistant operates autonomously to join and record meetings once configured, but acts under the ultimate supervision and setup of the human host.
• Novelty: Evolved capability. Represents an evolution of privacy risks where AI integration automates biometric data harvesting at scale during routine virtual communications.

Impact by dimension

• Physical security: Negligible. No physical security threats or critical infrastructure impacts were reported in connection with this commercial legal dispute.
• Information security: Negligible. The incident involves commercial transcription software rather than state-sponsored intelligence operations or information warfare.
• Sovereignty: Negligible. No state authority, border control, or core government operations were compromised or threatened by this commercial privacy dispute.
• Economic security: Negligible. While the company faces potential litigation-related financial losses, there is no threat to national economic stability or strategic tech sectors.
• Societal stability: Minor. The unauthorized collection of biometric voiceprints from non-consenting individuals represents a localized civil privacy concern, currently being addressed through standard legal channels.