Field Notes

source

The Online Safety Act

Are children safer online?

StatusStableUpdatedStands on1 sourceSupportsnotes

Internet Matters’ May 2026 report The Online Safety Act: Are children safer online? is the first substantial evaluation of the UK Online Safety Act from the perspective of families. The preserved report combines a survey of 1,270 UK children aged 9 to 16 and their parents, fielded by BMG Research from 15 September to 1 October 2025, with seven focus groups held in February 2026. Fieldwork began about seven weeks after the Act’s Protection of Children Codes came into force on 25 July 2025, so the report captures first impressions, not a settled outcome.

Findings on age checks

About half of children (53%) had recently been asked to verify their age, most often when setting up a new account. Where checks appeared, 37% involved facial age estimation, 24% a third-party application, and 22% a government identity document; a third of children (34%) abandoned the process when prompted. 84% of parents supported using age checks to keep children from pornography.

The checks are widely seen as weak. 46% of children considered them easy to bypass against 17% who called them difficult, and 32% admitted bypassing one within two months. The common methods were low-tech: a false birthdate (13%), someone else’s login (9%), or someone else’s device (8%). Only 7% had used a VPN, and children’s VPN use held stable at about 8% of 9- to 17-year-olds after the Act, which the report offers as a counterpoint to claims that age checks would drive children to VPN use. The government’s own commissioned study, Children’s circumvention behaviours online, later reached the same pattern with a larger sample: 7% of all children using a VPN to reach age-restricted services, with false birthdates the dominant bypass. More inventive attacks targeted facial estimation: submitted photos or videos of other people or fictional characters, and drawn-on facial hair, which one mother reported worked:

Estimation also failed honest users in both directions. Children misjudged as older kept using the platform with the wrongly estimated age, and children correctly rejected as too young could often try again almost immediately. A quarter of parents (26%) had allowed their child to bypass a check, with 17% actively helping and 9% turning a blind eye, usually where they judged the activity safe.

Findings on harm and design

49% of children reported a harmful experience online in the past month, including violent content (12%), content promoting unrealistic body types (11% of those 13 and over), and hateful content (10% of those 13 and over), all categories the Codes prohibit. Focus groups described distress at algorithmically surfaced footage of the assassination of Charlie Kirk, and children reported approaches from strangers and bullying.

The concerns families ranked highest sit outside the Act: time online and AI-generated content. Internet Matters’ companion 2026 wellbeing survey found 59% of children staying up late on devices and 46% continuing to watch or play when no longer enjoying it, and children in focus groups described infinite scroll and streaks as engineered to hold them. Children also reported AI material that was hard to distinguish from real media, including a nude deepfake made of a classmate.

Policy preferences

62% of parents supported banning social media for under-16s in November 2025 polling, up from 44% in August 2024, yet many supporters doubted a ban would work in practice, citing VPN workarounds and displacement toward harder-to-regulate services. Stronger enforcement of the OSA, stricter age checks, and restricting harmful platform features were the most popular alternatives. Only 22% of parents and 31% of children believed the government was doing enough. The report recommends safety-by-design, risk-based access rather than blanket bans, graduated age-appropriate experiences, highly effective age assurance, media literacy, and robust enforcement.

Limitations

The survey is cross-sectional self-report with no pre-OSA baseline, and the recall window for “recent” experience reaches back before the Codes came into force, so it cannot measure change caused by the Act. Internet Matters is a child-safety advocacy organization arguing for stronger regulation; its sampling and weighting are professional, but its framing and recommendations reflect that mission, and several cited figures come from its own companion reports.

The report grounds UK Online Safety Act, informs Age assurance, Age-verification circumvention services, and Less harmful social media, and supplies early evidence for Does age verification reduce harms other measures cannot.

Built on 1 source (1 external).

Working out connections…