The UK's communications regulator Ofcom will in January come out with requirements that online platforms will have to meet in order to enforce age verification, expected to be "highly accurate."

Judging by the statements made now by Ofcom's Online Safety Policy Director Jon Higham, the method will involve some type of facial recognition.

The result could be the blocking of "millions" of minors, while this is happening as a consequence of the country's controversial online censorship law, the Online Safety Act.

Previously, Secretary of State for Science, Innovation and Technology Peter Kyle said that unless tech firms "improve" age verification, the UK could move to ban users younger than 16 from social media.

Higham said in an interview that companies behind the platforms will have to implement technology that is highly accurate and effective and mentioned facial age estimation as being "pretty good" at guessing if a user is a child or an adult.

The report didn't go into how the planned age checks would work, but firms like Yoti that are selling this kind of service have users take selfies - this can be done in real-time with the company's own tool or the image can be sent to its API.

From there on, "AI" is used to analyze a face, assess age, and perform a liveness check. This is touted as a better (i.e., less privacy-invasive way) of ascertaining a user's age than having to present a government-issued ID to social sites.

However, users must trust those performing age estimation based on biometrics not to store the images of their faces and instead delete them immediately.

Tech companies that are found in violation of the Online Safety Act's rules will pay massive fines - up to 10 percent of global sales revenue, and if those infractions continue, their executives could spend up to two years in prison.