Meta Updates Tactics to Counter Foreign Influence Operations

Meta's latest 'Adversarial Threat Report' highlights various coordinated manipulation efforts detected and removed from Meta's apps in Q1 2023. Key findings include:

• Removal of 37 Facebook accounts, 13 Pages, five Groups, and nine Instagram accounts for coordinated inauthentic behavior originating in China targeting the global Sikh community.
• Disruption of operations from Bangladesh, Croatia, Iran, and Israel.
• Ongoing combat against the Russian network 'Doppelganger,' which aims to weaken international support for Ukraine.
• Observations on the use of AI in misinformation and deception efforts, noting that current AI-driven tactics have not significantly impeded Meta's ability to disrupt adversarial networks.

Meta has provided an update on the use of AI in misinformation and deception efforts, stating:

• Current Observations: Instances of photo and image creation, AI-generated video news readers, and text generation have been observed.
• Effectiveness of Countermeasures: So far, novel GenAI-driven tactics have not impeded Meta's ability to disrupt adversarial networks.
• Future Expectations: The expectation is that more groups will employ AI-based tactics over time.
• Existing Defenses: Meta's focus on behavior (rather than content) in countering adversarial threat activity appears effective, and the industry’s existing defenses are holding up well.

TikTok is under scrutiny due to concerns about its potential use as a vector for influence operations, given its Chinese ownership and significant influence outside of China. Meta's report suggests that Chinese-based groups have been actively seeking to use social media platforms to influence opinion on issues related to China’s political ambitions. This raises logical concerns that TikTok could pose a risk now or in the future, especially in light of findings that highlight the scale and ambition of such groups.

The 'Doppelganger' operation is a Russian network focused on weakening international support for Ukraine. It involves a large network of websites spoofing legitimate news outlets. Meta was the first technology company to publicly report on this operation nearly two years ago. The EU Disinfo Lab and the Digital Forensic Research Lab published open-source research at the same time. In December 2022, Meta attributed the operation to two companies in Russia, which were sanctioned by the EU in 2023 and by the US Treasury Department in 2024. Meta continues to combat this network by evolving its detection and removal approaches.

Meta is continuously evolving its approaches to detect and remove foreign influence operations before they can have significant influence. Key steps include:

• Monitoring and Assessment: Constantly monitoring and assessing the risks associated with evolving new technologies like AI.
• Behavioral Focus: Focusing on behavior rather than content in countering adversarial threat activity.
• Refinement of Detection Systems: Refining and revising detection systems to keep up with new tactics employed by threat actors.
• Persistent Efforts: Recognizing that foreign adversaries are always seeking to use high reach and high influence surfaces like social media to expand their messaging, and thus, maintaining persistent efforts to improve detection and removal capabilities.