Meta has outlined the latest developments in its evolving efforts to combat platform manipulation and hate speech, which have resulted in two significant network take-downs of late.
As outlined in Meta’s latest “Adversarial Threat Report”, Meta was recently able to remove two of the largest-known covert influence operations in the world, through a collaborative effort that could also help to chart a new way forward for future enforcement.
The two groups originated from China and Russia, and involved programs that targeted over 50 social media apps, including Meta’s tools.
- The Chinese operation has been labeled ‘Spamoflage’ in the cybersecurity community, and involved a complex web of programmatic efforts to manipulate Western news media by seeding positive commentary about China and the CCP. The initiative also looked to attack Western policies, and even specific journalists and researchers that have been critical of the Chinese Government. The initiative spanned across thousands of accounts and pages
- The Russian operation, meanwhile, entailed thousands of malicious website domains, which had each been running stories that mimicked the websites of mainstream news outlets and government entities, and posted fake articles, that had been largely aimed at weakening support for Ukraine. The program targeted users in France, Germany, Ukraine, the U.S. and Israel.
Meta says that these huge operations, running across many social platforms and websites, had been live for some time, meaning that this latest takedown, which may also lead to criminal prosecution in their respective states, could make a big dent in the influence operations space.
It’s a significant step, and Meta’s praised the broader collaborative approach that’s led to this breakthrough, which it’s hoping will also function as a disincentive to other bad actors in future.
In addition to this Meta’s also published a new study of the effects of six network disruptions of banned hate-based organizations on Facebook.
“The research found de-platforming these entities through network disruptions can help make the ecosystem less hospitable for designated dangerous organizations. While people closest to the core audience of these hate groups exhibit signs of backlash in the short-term, evidence suggests they reduce their engagement with the network and with hateful content over time. It also suggests that our strategies can reduce the ability of hate organizations to successfully operate online.”
This is also a major step, as it points to more effective approaches in combating the spread of hate speech online.
The network effects of social media help to connect users with like-minded folk, no matter where they are, which can obviously be a positive, but it also means that hate groups can amplify their message, and recruit more members, through the same process.
As such, establishing better ways to mitigate such could be a huge step, and this research could provide new guidance on this front.
Meta’s also outlined its approach to combating influence operations on Threads, and how it’s building this into the new app’s foundations, while it’s also shared new insights into how it’s looking to tackle misuse of its generative AI tools, through collaboration with researchers to seek out vulnerabilities.
Through live “stress tests”, Meta’s hoping to establish better ways to tackle these key challenges, which are already driving better results through expanded collaboration.
You can read Meta’s latest Threat Report overview here.