TikTok Opens a Brand Safety Center and Pushes AI Ad Labels
TikTok has launched a Brand Safety Center, a Brand Protection Playbook and a guide called AI Safety Essentials, aimed at advertisers who want to know what the platform does when their brand appears next to something ugly. The tools were announced at the company's APAC Responsible Media Forum in Singapore, which drew 30 representatives from brands, agencies and partners.
The pitch rests on a study TikTok commissioned from Ipsos across 12 markets. People who saw sponsored brand content on TikTok rated those brands 30 per cent more authentic and 11 per cent more trustworthy than people who had not seen it, according to the research, which used a test and control design.
And that is the commercial logic in a sentence. TikTok is arguing that authenticity is the product, so protecting the feed protects the advertising, and the new centre is the place where brands can see how that is being done rather than take it on trust.
Francis Stones, head of global responsible media and brand safety at TikTok, put the argument in those terms. "People come to TikTok for authenticity, and protecting that experience helps create an environment where brands can connect with audiences confidently," he said. "Our role is to build and maintain trust while giving brands the tools and safeguards they need to participate in culture with confidence."
The enforcement numbers behind the tools are the substantial part. Content goes through automated review first, with trained human moderators providing context, and potentially misleading claims are referred to more than 20 fact checking organisations accredited by the International Fact-Checking Network, covering 60 languages and 130 markets. In the second quarter of this year, TikTok removed more than 30 million videos across 10 Asia Pacific markets, about 0.3 per cent of everything published in the region, with 99.5 per cent taken down proactively and 91 per cent within a day.
And globally, automated systems accounted for 97.4 per cent of video removals in the quarter. That figure cuts both ways, since it shows scale and it shows how little of the work a human reviews before a video disappears.
But the advertiser tools go beyond removal. Brands get an inventory filter, category exclusion, vertical sensitivity settings and exclusion lists, with verification from outside partners including DoubleVerify, Integral Ad Science and Zefr. A new set of protection tools adds brand perception tracking in Business Center, reporting routes for copyright and trademark problems, and two systems for TikTok Shop, an intellectual property centre and a programme called TikTok Real that lets brands review seller authorisations and remove counterfeits.
The AI side is where the regulatory pressure sits. TikTok says it labels AI generated content through C2PA Content Credentials, creator labelling tools and invisible watermarking, and that it has now labelled more than three billion videos. "As more creators explore AI, we continue to expand labelling technologies and partnerships that help them do so responsibly," Stones said.
But for advertisers in this region the practical question is whether a dashboard changes behaviour. Brand safety tools only matter at the moment a campaign is paused or a placement is refused, and the centre is most useful if it makes those decisions visible after the fact.
TikTok launched a Brand Safety Center, a Brand Protection Playbook and an AI Safety Essentials guide
It removed more than 30 million videos across 10 Asia Pacific markets in the second quarter, about 0.3 per cent of uploads
99.5 per cent of those removals were proactive and 91 per cent happened within a day
Outside verification comes from DoubleVerify, Integral Ad Science and Zefr
TikTok says it has labelled more than three billion videos as AI generated
Source: TikTok media release
