New AI app Sway.ly launches to help children take control of social media algorithms

A new app called Sway.ly launched Tuesday with the aim of giving parents and children greater control over the algorithmic feeds that deliver social media content, its developers said. The app uses artificial intelligence to identify and filter 36 types of material the company classifies as harmful, send alerts to parents and provide suggestions to young people on safer digital choices.

The launch follows a company survey of 1,400 nine- to 19-year-olds that found more than three-quarters of respondents believe social media is damaging their physical and emotional health and that 72 percent had seen "uncomfortable" content that made them feel upset, sad or angry. More than two-thirds said they had experienced either a physical or emotional symptom they attributed to social media use, including sore eyes and tiredness (reported by more than a quarter), sleep problems (20 percent), feelings of addiction to phones (17 percent) and symptoms of sadness, anxiety or depression (14 percent). Neurodivergent children reported higher rates of harm, with 35 percent saying they had experienced cyberbullying compared with 20 percent of their peers.

Sway.ly's founder and chief executive, Mike Bennett, a father and technology entrepreneur, said the persistent "slow burn" of algorithmic recommendations is the primary danger for young people rather than isolated extreme posts. "Children are being overexposed to a relentless feed of toxic narratives about violence, misogyny, beauty, self-harm and more," Bennett said. "This is not content that they go looking for — it's what finds them. It is delivered through memes and algorithmic recommendations, and it gradually reshapes their sense of self and the world around them."

The app, which the company said was created using peer-reviewed research and input from psychotherapists and AI specialists, continually monitors fast-changing youth culture, trending language and formats in order to stay current. When it detects content the company regards as harmful, Sway.ly can filter that content in a young person's feed, prompt the user with educational suggestions — such as unfollowing or blocking accounts that glamorize self-harm or unrealistic body images — and notify parents of recurring concerns.

Daniela Fernandez, Sway.ly's chief strategy officer, said parents increasingly struggle to keep up with rapidly shifting online trends and that blanket bans often lead to conflict. "Most parents simply can't keep up with the pace — the language, the trends, the sheer volume of content is constantly shifting," she said. "We wanted to build an app to keep up and help decode what kids are really seeing, and to give parents the insight and tools they need to respond with confidence. It's about rebuilding trust between kids and parents — and using AI for good, to support smarter habits and healthier minds."

Psychotherapist and online-harms consultant Dr. Catherine Knibbs said removing harmful material entirely is impractical because such content is varied and adaptive, and that education and empowerment are essential. "When we create safe spaces for children to talk about what they see online, we empower them to cope, reflect and choose," she said.

Sway.ly is offered as a subscription service at £2.60 per user per month, the company said. Bennett and Fernandez argue the tool is designed to complement, rather than replace, parental involvement and public policy. They welcomed regulatory efforts such as the Online Safety Act but said legislation that focuses mainly on bans can miss the subtler, persistent harms produced by recommendation systems.

Experts and advocates have increasingly warned about the algorithmic "drip" that normalizes unrealistic body standards, violent imagery and other forms of content that can undermine young people's well-being. Parents often express concern about extreme or explicit posts, but Bennett said the research shows the cumulative exposure to curated, benign-looking material can be equally damaging over time.

Sway.ly's developers said the product tracks evolving youth digital culture, including coded language and emerging formats, to keep filters and guidance up to date in real time. They said the intent is to give families tools for gradual digital education and to reduce confrontation that can arise from blanket bans or unilateral restrictions.

The company did not publish independent validation of the app's content-classification accuracy at launch. The broader debate over how best to protect children online continues to involve a mix of technological tools, parental education, platform responsibility and public policy.

Sway.ly is available starting Tuesday through its website and app stores, the company said, and will be marketed to parents, schools and mental-health professionals as part of an effort to combine technology, research and family-based education to reduce online harm among young users.