A Reddit thread is raising eyebrows with claims about Google’s Call Screen feature. According to a post by user Jay_Kane123, the virtual assistant now starts calls with a notice saying the conversation “may be used to improve Google AI.” The idea of recorded calls being used for AI training has sparked concern. People are wondering if this step is a slippery slope toward more data collection disguised as convenience.
The claim hasn’t been confirmed by Google. There’s no audio recording or official announcement to back it up. But the post gained traction quickly, with hundreds of upvotes and a heated discussion about privacy and practicality. Many users shared their frustrations. One wrote that most callers hang up right away when they hear the assistant’s introduction. Another complained that legitimate callers get scared off by the mention of recording.
This isn’t even surprising. If I heard an automated voice on the other end, I’d probably hang up too. However, for businesses where hanging up on the customer isn’t the best step, hearing something like “may be used to improve Google AI,” would most likely be a red flag and could result in them hanging up the call.
What’s frustrating is that it’s not the first time Google’s efforts to gather data for AI training have raised concerns. Last year, reports surfaced about Google quietly updating its privacy policy. The fine print mentioned that everything from Google Search queries to Gmail content and Google Drive files could be used to train its machine learning models if that data is publicly available. That announcement didn’t sit well with privacy advocates, who warned that Google might be getting too comfortable with mining user data.
Then, there’s a whole other deal that Google struck with Reddit to feed as much data it can to train its AI on natural human interactions. So it’s clear that the Call Screen feature’s latest change feels like part of a bigger pattern. It’s one more example of how tech companies are using everyday interactions to train their algorithms. And while Google frames these updates as helpful innovations, not everyone is buying it.
Some Pixel users on the thread argue that training AI with call data makes the feature better at spotting spam calls and improving transcriptions. But even those who like the tech admit it needs work. Many callers still don’t respond to the AI prompt, leaving users frustrated with awkward pauses instead of useful call screening.
Without solid proof, it’s unclear if Google has officially added this AI training clause to its call screening feature. Still, the lack of pushback from other users suggests it might be real — or at least believable enough to make people worry. Could similar AI disclosures soon come to other Google services? Will this lead to even more data collection practices hidden in plain sight? I guess only time will tell.
In the meantime, feel free to share your thoughts on the Call Screen feature apparently telling callers it’ll use the recording to train its AI.