TL;DR
O2 said on 30 October that it has moved from warning customers about suspected scam calls to blocking them outright, using AI-assisted technology from Hiya, at no extra cost. The company expects to stop more than 3 million calls a week. The figures and the “first UK operator” claim are its own, but a network-level decision to block rather than warn is a real step, and one that places the judgement with an algorithm.
Nothing needs installing, as the service reaches compatible devices automatically. O2 says it has flagged 1.6 billion-plus calls as suspected scams or spam since the start of 2024, and that this history gave it the confidence to act before the phone rings. Suspected spam will still only be flagged.
What the trial suggested
According to the company, scammers often failed to notice the block. In testing, 60% of blocked calls were repeat attempts made inside a minute, and on average a number labelled as fraudulent tried three times as many calls after being blocked as before. O2 presents this as disrupting fraudsters. It may also show how persistent these operations are.
Fraud Minister Lord Hanson welcomed the move, linking it to the government’s fraud strategy and its telecoms charter with mobile networks. O2 says no other major UK network includes advanced AI flagging and blocking free, though it acknowledges that rivals have scam protection of their own. We have not seen those claims independently checked.
The question the release leaves open
What O2 does not publish is an error rate. It says legitimate calls will continue to get through, but gives no figure for genuine calls wrongly stopped, and the release does not describe how someone whose number is blocked in error can get it corrected. In our view, this matters more once a system moves from advising to deciding. A false warning can be ignored. A blocked call simply never arrives, and the person affected may not know why, whether that is a GP’s surgery or a small business with a newly assigned number.
The wider context
The company’s own Scam Index gives a sense of the pressure. It says 61% of Britons who encountered a suspected scam between July and September received a suspicious phone call, up five percentage points on the previous quarter, while the release also points to an Information Commissioner’s Office investigation into robocalls. Hiya’s own research says 31% of respondents got a deepfake over the past year, a figure that comes from a company selling call protection.
Looking Forward
Other operators now have a comparison point on both protection and pricing. In our view, the next useful disclosure from O2 would be how many genuine calls were stopped, and how quickly they were reinstated, because that is the evidence regulators and customers will need to judge the trade-off. Customers can still report suspicious calls and texts free to 7726.