Online platforms will have to ban scam ads under UK proposals to tackle fraud

Range of potential measures announced by Ofcom include reducing risk of accounts being hijacked for scams

Big tech platforms will be required to ban scam advertisers in the UK under proposals to tackle online fraud.

Facebook, Instagram, Snapchat, X and YouTube will have to block bad actors who post fraudulent ads and prevent them from creating new accounts in measures targeted at the biggest services.

The potential changes include: reducing the risk of accounts being hijacked and turned into hosts for scams, ensuring that ads for banking or financial services have legal clearance, and giving law enforcement channels for identifying fraudulent ads.

Ofcom, the UK’s communications regulator, announced the proposals under the next phase of implementing the Online Safety Act (OSA), which places responsibilities on tech companies to protect users from harmful content.

The scam ad requirements apply to large social media platforms, as well as Google and ChatGPT. Fraudulent ads cost victims in the UK about £200m a year, Ofcom said.

Oliver Griffiths, Ofcom’s online safety group director, said tech groups had not done enough to combat fraudsters on their platforms. “We expect firms to take robust action to stamp out scam ads and boot out the bad actors behind them to safeguard their users,” he said.

Once the new measures come into effect and become legally binding under the OSA, failure to implement them could result in fines of up to 10% of a platform’s global revenue. Ofcom urged firms to improve their anti-fraud measures immediately, but the consultation on specific steps that launched on Friday closes in October and final decisions will not be made until next year.

The consumer group Which? welcomed the proposals as it accused tech firms of treating scam ads as a “profitable income stream”, but expressed concerns that measures would not be implemented until next year.

Rocio Concha, the head of policy and advocacy at Which?, said: “This is very problematic at a time when breakneck advances in AI are making scams more sophisticated than ever.”

The Bank of England warned the public last month against falling for AI-generated scams after deepfake videos of Nigel Farage fighting its governor spread online . Martin Lewis, the finance expert, has repeatedly urged the government to tackle scam ads that use his image .

The advertising announcement came as part of a wider consultation on measures under the OSA for “category 1” tech companies, which are Facebook, Instagram, Pinterest, Quora, Reddit, Roblox, Snapchat, TikTok, WhatsApp, X and YouTube. Google and ChatGPT were assigned to a secondary category that avoids some of the restrictions but requires them to adhere to the scam ad measures.

As part of the consultation, Ofcom also set out draft rules for category 1 companies on the treatment of journalistic content. The proposals include: ensuring that social media platforms do not arbitrarily restrict access to news content and “content of democratic importance”; giving news publishers the opportunity to state their case before content is removed, labelled or down-ranked; and offering an expedited complaints process for journalistic content.

The watchdog published proposals for category 1 platforms to limit user exposure to content that features suicide, self-harm, eating disorders, or hate and abuse. Users must be given the option to reduce their exposure to this content, Ofcom said, which could include removing it entirely or blurring it. Users must also be given the option to block or mute other users and filter out interactions with non-verified accounts.

Ofcom also set out category 1 platform content rules, which explain what sort of material is allowed on a service and how to implement those terms consistently.