Our newest Adverts Security Report highlights a key pattern from 2024: how AI is bettering our skill to forestall fraudsters from ever exhibiting advertisements to individuals.
For years, we have deployed our most superior applied sciences to safeguard our advertisements platforms from dangerous actors. In 2024, we launched over 50 enhancements to our LLMs, which enabled extra environment friendly and exact enforcement at scale. These updates sped up complicated investigations, serving to us determine dangerous actors and fraud alerts — like illegitimate cost data — throughout account setup. This stored billions of policy-violating advertisements from ever exhibiting to a client, whereas guaranteeing professional companies can present advertisements to clients quicker.
We additionally tailored our defenses towards ever-evolving scams, notably the rise of AI-generated public determine impersonation advertisements. To combat again, we assembled a group of over 100 specialists to develop countermeasures, akin to updating our Misrepresentation coverage to droop advertisers that promote these scams. Consequently, we have been capable of completely droop greater than 700,000 offending advertiser accounts. This led to a 90% drop in reviews of this type of rip-off advert final 12 months.
That is only a fraction of our ongoing work to maintain the promoting ecosystem secure. Study extra in our 2024 Adverts Security Report.
