Our children deserve a SAFE, not safer internet

Since its inception way back in 2004, Safer Internet Day has been about education, guidance, support and access to resources to help children explore the online world safely. This year’s theme is ‘Smart tech, safe choices – Exploring the safe and responsible use of AI’. But this year, more than any year, the day and its objectives seem like a throwback to a more optimistic age. Our children are still not safe online and, with the rise of AI tools and chatbots, they are becoming less and less safe everyday. We need urgent action to expand, strengthen, and enforce laws such as the Online Safety Act to protect our children immediately.

In the wake of the scandal with X’s chatbot Grok being used to create illegal non-consensual intimate images, the US court cases seeking justice for vulnerable young people who killed themselves after interacting with chatbots, the alarming “Moltbook” platform for AI chatbots, and other daily stories about risks and harms, suggesting that there is a possibility for children to experience the “safe and responsible use of AI” seems rather quaint.

AI developers are rolling out products with a speed that is downright reckless, with an eye on their profits rather than a concern for their safety. Regulators - like Ofcom - seem impotent in the face of the scale of harm occurring on their watch and blame the legislative tools they were given. Parents and Parliamentarians are losing patience, and discussions and debates about online safety are now dominated by calls for social media bans - in the UK and throughout Europe.

But we can’t ban or educate our way out of a problem that is fundamentally about the rollout of unsafe, untested products at scale. We have to call on Governments and legislators to get ahead of the problem and use the well-established, proven levers at their disposal: the law and strong, enforceable regulation.

Successive Governments have been reluctant to address the need for AI-enabled products or services to be treated - and regulated - like any other such product and service on the market: nearly 50 MPs wrote last week to the Technology Secretary, in a letter coordinated by Labour MP Anneliese Dodds, flagging the fact that Grok was clearly launched without adequate safety testing and risk assessment, or a focus on “safety by design” - all things that we have called for the Government to address via our 10 point plan to strengthen the Online Safety Act.

Ofcom last week set out clearly that “not all chatbots are covered by the OSA” - again, we have been pointing this out for months and are currently developing our own proposals to fill that legislative gap. More broadly, we’d agree with Dodds and her colleagues that the Government must get on with the introduction of AI legislation to address the potential for widespread harms from the unsafe use of that technology.

Children are already using AI tools and applications - in schools, embedded in search and social media platforms, via chatbots and other AI-enabled products. Industry has shown no enthusiasm for developing it safely or responsibly and, for all the good intent of Safer Internet Day, it’s just not possible for education alone to keep children safe.

We are urging the Government to take simple immediate action that could make the internet safer for our children from tomorrow. They can strengthen the OSA to ensure no unsafe products come to market, all products are safe by design and new harms from AI tools and chatbots are included.

So, on Safer internet Day 2027, instead of sharing warm words and future commitments, we might all be able to celebrate how the OSA has made our children, and many, many other people safer.