Zuckerberg, please pay attention! Meta community scams have caused over 1.3 billion in financial losses, and the Executive Yuan demands an immediate change to the algorithm.

The Executive Yuan states that Meta platforms involved in fraud losses exceeding NT$1.3 billion in the first four months of this year, strongly urging them to modify their algorithms to block scam posts. If they respond passively, the government will consider legislative amendments to impose heavier penalties. Additionally, Meta faces legal challenges in the United States over immunity rights related to AI advertising.

Fraud rampant! Executive Yuan demands Meta communities immediately modify algorithms

The Fraud Prevention Command Center of the Executive Yuan announced on May 20 that Meta must proactively modify its algorithms and detection systems immediately to block repetitive scam posts at the source. If Meta continues to respond passively, amendments to the "Fraud Crime Prevention Act" will be proposed to introduce regulatory measures and penalties.

The Executive Yuan cited data from the Criminal Investigation Bureau indicating that from January 1 to April 30, 2026, the number of scam-related cases on Meta’s social platforms reached 10,4073, accounting for 84.31% of cases, with victims losing over NT$1.3 billion, representing 86.2% of the financial losses.

These figures demonstrate that Meta’s social platforms remain the primary channels used by scam groups. Common scam posts often exploit seasonal or current events, impersonate government agencies to sell agricultural products, or claim store closures to sell electronics at low prices, often using photos of physical storefronts to gain victims’ trust.

Image source: Complete flowchart of AI and current event-based scams by scam groups on the Executive Yuan

The Executive Yuan proposes five major anti-scam measures, hoping for active cooperation with Meta

To effectively curb scam posts, the Anti-Fraud Center has proposed five anti-scam strategies. Including Meta proactively fixing technical vulnerabilities and modifying algorithms, establishing mechanisms to identify repeated scam content, and intercepting information already flagged as scams; calling on Meta to provide trustworthy identity verification services for local small businesses to help consumers distinguish legitimate businesses from scam accounts, among others.

The Executive Yuan does not rule out legislative amendments to strictly regulate, and will evaluate future penalties based on the number of victims or scam posts, imposing heavier fines, and supports legal remedies, encouraging victims to seek damages from platforms.

According to "Public Television News Network," Deputy Director Hou Yifang stated that Meta has joined the anti-scam efforts, having proactively removed 7.8 million scam posts last year; Director Lin Junxiu of the Industry Department of the Digital Development Agency also mentioned that they will continue to work with Meta to introduce more AI technology, striving to reduce the reach of scam content through technical means.

U.S. court rules Meta AI advertising cannot be exempt from liability

Last year, Reuters reported that internal Meta documents estimated that about 10% of total revenue in 2024 would come from scam and illegal ad placements, amounting to approximately $16 billion.

The documents further indicated that Meta’s automated systems only prohibit ads when there is at least a 95% certainty that the advertiser is a scam; if the probability is lower, higher ad rates are charged. However, Meta spokesperson Andy Stone countered that this figure was a rough estimate, and the actual number is lower, emphasizing that reports of scam ads from global users have decreased by 58%.

• Related report:** Zuckerberg making easy money? Reuters: 10% of Meta’s revenue from scam ads, 99% of scam media in Taiwan originate from Meta**

In May of this year, a U.S. federal court made a significant ruling regarding a cryptocurrency scam case, stating that Meta’s AI advertising tools may have exceeded the immunity protections under Section 230 of the Communications Decency Act, as the plaintiff accused scammers of using Meta’s AI system to disseminate false investment information on a large scale.

The court held that, if platforms actively assist in ad optimization and target audience recommendations through AI systems, they are involved in the promotion process of scam content and cannot be considered entirely neutral platforms. This ruling is expected to impact compliance policies of social media platforms and may influence AI advertising and content recommendation models of major tech companies in the future.

• Related report:** AI advertising tools fueling scams! Meta ruled liable, may face securities fraud charges**