Allegations of Unauthorized Recording of Private Conversations
A recent lawsuit has been filed against a tech company known for its transcription services, alleging that the company has been secretly recording private conversations. This information, it is claimed, is then used to train the company's artificial intelligence (AI) systems without the consent of the users involved.
The company's transcription service, which is known for providing real-time transcriptions of online meetings, is accused of not requesting permission from meeting participants to record. Furthermore, the lawsuit alleges that the company does not notify participants that these recordings are being used to enhance their AI systems.
The Plaintiff's Case
The plaintiff, a man from California, claims that his privacy was grossly violated when he discovered that a confidential conversation he was part of was being secretly recorded. Filed in a district court in California, the lawsuit alleges that these covert recordings go against both state and federal privacy and wiretap laws. The lawsuit is also seeking to represent others in California who've unknowingly had their chats shared with the company, which the plaintiff's lawyers argue is a tactic for financial gain.
Though requests for comment have been made, neither the legal team representing the plaintiff nor a spokesperson for the tech company have responded.
The Privacy Policy: Explicit Permission or Deception?
The tech company's privacy policy does not veil the fact that it uses AI training. It states that it receives "explicit permission" from users to train its systems using meeting transcripts when users tick a box allowing the company and third parties to use private conversations for "training and product improvement purposes". However, the lawsuit argues that many users are still being deceived.
In the past few months, the company has faced growing privacy concerns as it became more widely used in workplaces around the globe. The company claims that approximately 25 million people are now using its AI transcription tools, which have recorded and processed over 1 billion meetings since the company was founded.
User Complaints and Concerns
Users have voiced their concerns on various platforms about the company's automated recording tools. For instance, an AI researcher and engineer reported that a meeting he had with investors was recorded without his knowledge. The transcript of the meeting, which included confidential business details discussed after he left the meeting, was shared with him. This resulted in a business deal falling through.
Additionally, a foreign correspondent reported on an interview with a human rights activist that was conducted using the transcription service. The correspondent expressed fear about the company sharing user data with third parties, which could potentially allow foreign governments to gain access to raw transcriptions of sensitive conversations.
Further complaints have been made about the service joining meetings automatically when linked to workplace calendars, and recording chats without consent.
The lawsuit also points out that if a user has an account with the company and joins a virtual meeting, the software will generally ask the meeting's host for permission to record. However, it does not typically ask permission from all other participants.
De-Identification: A Solution or A Problem?
The company claims that before the audio of meetings is fed into its machine learning systems to improve its speech recognition feature, it is "de-identified". This process is supposed to anonymize the data. However, the lawsuit questions the company's ability to do this effectively, noting that the company provides no public explanation of its "de-identifying" process.
The lawsuit argues that the company's de-identification process does not ensure the removal of confidential information or guarantee speaker anonymity.