Antigone Davis arrives in Spain amidst a political storm she has been dealing with for several years. As Meta's Safety Director, she is the voice of the social media platform in one of the most debated issues in technology today: the control of minors on social networks. This issue divides Europe between controls or prohibition, a path that Australia has decided to take.
Yesterday, the archipelago became the first country to prohibit its minors under 16 from accessing TikTok or Instagram and forced these companies to delete hundreds of thousands of teenage accounts. In an exclusive interview with EL MUNDO, Davis addresses the problems of this approach and shows her alignment with the Spanish government's vision, despite the summons from the PSOE in Congress to the top executives of her company a few days before her arrival in the country.
What is the reason for your visit to Spain?
I am visiting several countries in Europe. In Spain, we are meeting with government officials to discuss our approach to young people. Spain is interesting because it is very focused on parental controls, something that we strongly support and want to see implemented. It makes a lot of sense for families.
And what method do you advocate for?
What we would like to see is some form of age verification in the operating system or app store when it comes to minors. It could even be that the responsible adult who purchases the phone attests to the age so that developers can offer an appropriate experience for that age. It's a simple way to know who the user's parents are. Then, it could lead to a situation where if you are under 16, your parents have to approve the download of an application.
A year ago, you launched your teenage accounts along those lines.
With this account, if you are under 18, you have a private account by default, message restrictions, no notifications at night, and if you are under 16, you cannot change this. This provides peace of mind, but some parents want to go further: perhaps removing all notifications or blocking Instagram at night. We want to ensure that we can provide them with the possibility and access to those controls. There is a debate on this at the European level, but in Australia, they have just decided to ban social networks for those under 16.
There is a debate on this at the European level, but in Australia, they have just decided to ban social networks for those under 16.
We are complying with the law in Australia. The challenge with a ban is that no one wants to prohibit teenagers from accessing all applications, so you end up banning a small number of them. When you do that, kids look for other apps outside the ban that regulators are not aware of, like Lemonade. There are between 1.5 and 2 million apps in the stores, and there is no way a regulator can chase teenagers from one to another. I understand the motivation behind these bans, but I don't think they will achieve what they aim for. Beyond age verification, we believe that social networks also need to have obligations to show age-appropriate content, as we have done by drawing inspiration from the PG rating in movies, and allow for setting time limits for usage.
Is verifying the age of minors a challenge?
Yes. Initially, we ask for age, but we also have technology. We are very good at determining if someone is over or under 18. It works very well, but it is more complicated to determine the age between 13 and 18, where the accuracy drops to 34%. Sometimes these children do not have documentation. That's why age verification after buying the phone and having to create an account is a moment shared by the adult and the minor, allowing for a more precise verification. And it respects privacy since they only have to say if they are under 18 or under 16 years old.
The Spanish government has called several Meta executives to appear in Congress. Do you intend to attend? What do you think of the President of the Government saying that they spy on Spaniards?
Meta takes privacy very seriously and offers various tools to help people control the use of their data. We hope to collaborate constructively with the authorities on this matter.
P. Meta is entering new businesses such as artificial intelligence (AI) and glasses. Precisely, AI and minors are another hot topic.
Yes. The key for regulators is to build laws that are resilient to the rapid advancement of technology. Now, the focus is on how young people interact with AI chatbots. If you think about the parental consent model in the App Store or app store, it is a resilient model. The gateway to technology is always through an app that is downloaded. It is an initial line of defense that can be placed there.
AI also has other risks.
People are concerned about risks like image recognition generated by AI. Companies have become quite good at this and are already doing it relatively well. But AI also offers good possibilities. I work on issues like child exploitation. One of the things we see is that people who commit these crimes use specific language that tries to avoid detection. They use different terms or put exclamation marks between words to hide them. AI is helping a lot to improve how we deal with this. We no longer need to find the keyword; we follow the communication pattern and it leads us to who is trying to exploit or harass a minor. I often tell regulators that it's a bit like the game Whac-A-Mole. You hit one, but another one pops up. AI can help us win.
We have talked a lot about regulation, but in the US, companies like X or the Government seem to be going in the opposite direction. Does this concern you?
I believe that deregulation is not being applied when it comes to minors; there is a common thread that recognizes that these individuals are more vulnerable. Then, there are significant differences in how each actor thinks about solving these problems; some are very focused on taking control away from parents or following Australia's lead in not worrying about it. Others are more aware of the importance of technology and are trying to find another way.