Private video calls, text messages and meetings on Zoom might be used to ‘train’ artificial intelligence models.
The San Jose company’s new terms and conditions – which came into force in March but were spotted this month – have sparked a wave of outrage online, with users threatening to cancel their accounts over the change.
In one section of the new T+C’s, it says that customers consent to Zoom using data for purposes such as ‘machine learning or artificial intelligence (including for the purposes of training and tuning of algorithms and models).’
Artificial intelligence models are commonly trained with large amounts of publicly available data, often taken from the internet – but Zoom’s move would use private customer data, raising privacy fears.
The changes have caused a privacy storm (Reuters)
The changes came in paragraph 10.4 of Zoom’s Terms and Conditions (Zoom)
Zoom has responded with a blog post this week, claiming that the data is only used to train AI models to summarize meetings more accurately, and only with customer consent.
In a blog post, Zoom’s Chief Product Officer Smita Hashim wrote: ‘To reiterate: we do not use audio, video, or chat content for training our models without customer consent.’
But that hasn’t stopped users from growing concerned about their privacy.
Elliot Higgins of news organization Bellingcat said: ‘We run our training workshops on Zoom, so Zoom is effectively planning to train its AI on our entire workshop content with no compensation, so bye-bye Zoom.’
The data used to ‘train’ AI models has become a legal battleground in the wake of the rapid advances in generative AI technology such as OpenAI’s ChatGPT.
Users and businesses fear that private information could be inadvertently revealed – or artistic works plagiarized – if their data is used to train AI models.
Google recently changed its privacy policy to allow the company to train AI models on publicly available information on the internet.
The new policy says, ‘For example, we use publicly available information to help train Google’s AI models and build products and features like Google Translate, Bard, and Cloud AI capabilities
High-profile users threatened to cancel Twitter accounts (X)
Users were outraged at the changes to the privacy rules (X)
Elon Musk revealed a plan to train his xAI model on publicly available content on Twitter, saying: ‘We will use the public tweets—obviously not anything private—for training…just like everyone else has.’
In response, artists threatened to delete their Twitter accounts, fearing that their work could be used to train AI systems which would then copy their art styles.
In April, Europe’s national privacy watchdogs set up a task force to address issues with ChatGPT after Italian regulator Garante had the service taken offline, accusing OpenAI of violating the EU’s GDPR, a privacy regime enacted in 2018.
ChatGPT was reinstated after OpenAI agreed to install age verification features and let European users block their information from being used to train the AI model.
Jake Hurfurt, head of research and investigations at privacy organisation Big Brother Watch told DailyMail.com, ‘Companies should be transparent about how they use their customer’s data.
‘Hiding huge changes in how they process user data in dense privacy policies is unacceptable.
‘People should have the power to choose how their sensitive data, including biometric data is processed, and should be given clear information and the ability to opt out.’
In Zoom’s blog post, Hashim explained that customer data would be used to train AI models used to summarise meetings, and customers have to consent to it first.
She wrote, ‘When you choose to enable Zoom IQ Meeting Summary or Zoom IQ Team Chat Compose, you will also be presented with a transparent consent process for training our AI models using your customer content.
‘Your content is used solely to improve the performance and accuracy of these AI services. And even if you chose to share your data, it will not be used for training of any third-party models.’