Meta expands anti-fraud protections in Poland, citing a surge in impersonation and scam advertising

Meta says it is stepping up protections against online scams in Poland, pointing to a rise in fraud across platforms, dating apps, online games, crypto-related schemes, and SMS-style messages. In a policy update published Monday, the company said it is responding to tactics that try to evade detection and exploit trust, including impersonation of celebrities and other public figures through ads and accounts designed to look legitimate.

The update frames the changes as both a user-safety effort and a business necessity. Meta said fraud hurts victims and also undermines trust in the advertising ecosystem that underpins its revenue model. The company also disputed what it described as a recent anti-Meta campaign that “distorts” its motives and approach, arguing it relies on inaccurate reporting and selectively quoted, leaked documents rather than on what Meta says it did after assessing the scale of the problem.

Meta said it already relies on advanced artificial intelligence systems to stop more scams and to identify and block more accounts impersonating celebrities and other public figures. The company attributed earlier progress in Poland to these tools, saying they have reduced by more than 80% the number of user reports related to the most commonly used images of known people.

To address evolving scam tactics, Meta said it is launching an upgraded AI system in Poland designed specifically to further limit impersonation cases. The company said Poland is in the first group of European countries where the new system is being deployed.

On the advertising side, Meta said it currently requires advertisers to confirm their identity in certain circumstances, such as when an account shows potentially suspicious behavior. Because financial services are a category frequently exploited by scammers, Meta said it is extending that identity confirmation process to advertisers of financial services targeting Poland. Meta added that the rollout will complete in the coming weeks.

Meta also offered targets for broader verification across its ad business. It said it aims for 90% of Meta’s global ad revenue to come from verified advertisers by the end of 2026, compared with 70% in 2025. The company said it will continue investing in verification processes to stay ahead of scammers’ methods.

In support of its claims, Meta published performance figures for Poland. It said that between July 2024 and June 2026, the rate of reports involving ads containing fraud fell by 83%. Separately, it said that from July 2025 to June 2026 it removed 137,000 ads characterized as fraudulent in Poland, with more than 88% removed before anyone reported them.

Meta described its anti-fraud effort as multi-layered. It said it protects users with automated systems, works to disrupt scam networks, collaborates with public institutions and private-sector partners, and provides tools intended to help people recognize and avoid scams, including warnings and safety features available on its platforms.

The company acknowledged that its choices are subject to public scrutiny, and said it values constructive criticism when it is grounded in facts. However, Meta did not disclose details in the update about the specific technical design of the “improved” AI system in Poland, the exact thresholds that trigger identity checks, or the extent to which the AI changes affect false positives, user appeal processes, or advertiser onboarding timelines.

As the rollout progresses, Meta’s next steps will likely be closely watched by regulators, advertisers, and consumer-rights groups, especially given the company’s emphasis on financial-services identity verification and the reported reductions in fraud-related ad issues. Observers may also look for how Meta measures the upgraded system’s performance after deployment, and whether similar identity-check expansions appear in other European markets.

Meta Newsroom said its actions underscore a continuing commitment to fight scams, but the update leaves some questions open about which scam categories are declining fastest and how the company will adapt if adversaries shift to new channels or creative styles. The company’s public challenge to a broader anti-Meta narrative also suggests it expects the debate to continue as more data and assessments emerge.