In April and May, a strange phenomenon struck the social media platform Bluesky. Hundreds of everyday users suddenly discovered that their accounts had been hijacked, locked out, and weaponized to flood the network with content about Ukraine. Months later, researchers at Clemson University unraveled the breach, attributing the operation to the Social Design Agency (SDA). The Russian consulting firm is heavily sanctioned by the U.S. Department of Treasury for running Kremlin-directed malign influence campaigns. Yet, as investigators soon realized, the Bluesky account takeovers were merely a minor node in a vast, sprawling portfolio of recent operations. Leaked internal documents obtained by the news portal Delfi Estonia and shared with the Organized Crime and Reporting Project (OCCRP) exposed far more alarming intentions: plans for real-world acts of vandalism, coordinated intimidation, and direct interference in foreign elections.
For years, public discussions and policy debates surrounding online disinformation have operated under a flawed assumption—that attacks like the one on Bluesky are isolated, discrete efforts. However, the operational blueprint of the SDA demonstrates that modern foreign influence campaigns are deeply interconnected, forming multi-layered propaganda strategies that completely transcend traditional boundaries. Sophisticated state actors, including China, Iran, and Russia, routinely combine overt public messaging with covert underground influence, seamlessly blending traditional broadcast channels with decentralized social media platforms. State-backed media outlets, official government accounts, paid online influencers, fake personas, hacked profiles, and automated bot armies are no longer separate entities; they are synchronized cogs in a singular, monolithic system.
This integrated methodology, which experts term "full-spectrum propaganda," has rapidly become the operational norm for adversaries attempting to project power and sway public opinion abroad. As outlined in a recent analysis published in the academic journal Security Studies, full-spectrum propaganda relies on multiple reinforcing channels that validate one another, making fabricated narratives appear exceptionally credible and allowing them to proliferate at unprecedented speeds. A story publicized on a state-backed broadcast network may successfully reach one demographic, while covert social media accounts capture the attention of an entirely different digital audience. Official government outlets can quote posts from entirely fake accounts whose operators masquerade as ordinary members of the public, thereby normalizing fringe claims. Meanwhile, paid influencers tailor messages to resonate authentically with niche communities. The rapid, unchecked expansion of artificial intelligence is supercharging this strategy, allowing small groups or even solitary actors to conduct massive, complex propaganda campaigns that would have previously required state-level resources and vast bureaucracies.
Strategies that merely address discrete, individual foreign influence campaigns are fundamentally unequipped to meet the multifaceted challenge of full-spectrum propaganda. The United States desperately needs the capability to assess, understand, and counter these influence operations holistically. Yet, paradoxically, as the threat continues to scale exponentially, the United States has actively dismantled the institutional capacity required to do so. Political pressure and policy shifts have led to the systematic dissolution of government agencies designed to track foreign interference, while social media companies have simultaneously erected high barriers preventing researchers from analyzing their platforms. As AI accelerates the velocity and broadens the global reach of foreign propaganda, Washington is effectively blinding itself to the challenge.
EVERYTHING EVERYWHERE ALL AT ONCE
For much of the past decade, politicians, regulators, and the general public understood foreign influence operations on social media strictly through the isolated fragments that first became visible. Most commonly, these discoveries took the form of a network of fake accounts unearthed by social media platforms as part of routine enforcement efforts to eliminate bots, scams, and digital deception. Following the removal of such accounts, companies would publish transparency reports, and external researchers—including the authors of the Security Studies paper—would publish detailed analyses dissecting the evolving tactics, techniques, and procedures of online threat actors. This granular focus on the minutiae of individual operations conditioned policymakers and the public to view these threats as narrow, disconnected incidents.
However, viewing only the visible fragment misses the broader reality. State-backed propagandists do not operate under an exclusionary choice; they do not have to select between state media and fake accounts, between covert personas and overt broadcasters, or between spoofed media outlets and encrypted Telegram channels. They utilize all of them simultaneously—and with increasing frequency.
The Social Design Agency, for instance, documented its own covert work in the meticulous style of a corporate consultancy reporting directly to a paying client. According to leaked internal communications, that client was the administration of Russian President Vladimir Putin, whose senior officials routinely oversaw the agency’s operations. The consulting firm’s covert digital actions are simply one component of a sprawling Russian propaganda apparatus that simultaneously deploys state-owned media networks, official diplomatic channels, and fabricated online personas. Moscow and its proxies deploy these diverse tools in tandem to amplify core propaganda narratives and penetrate varied demographic audiences.
As the Organized Crime and Reporting Project and other investigative outlets revealed, the SDA was behind a sophisticated fake video released in June 2025. Branded meticulously to mimic the legitimate Saudi media outlet Al Arabiya, the video falsely claimed that Ukrainian President Volodymyr Zelensky had purchased a luxury apartment inside the Burj Khalifa skyscraper in Dubai. An internal SDA report estimated that this single fabrication could have accumulated up to 86 million views globally. A similar SDA effort targeting Armenian Prime Minister Nikol Pashinyan likely generated over ten million views. In other instances, rather than impersonating an existing real-world broadcaster, the SDA invents entirely fictitious news outlets from scratch. One leaked document detailed plans to deploy a shell outlet tied to the SDA during Armenia’s parliamentary elections.
When a greater degree of distance and plausible deniability is required, the SDA relies on paid social media influencers. Earlier this year, for example, the Kremlin sought to bolster the reelection campaign of Viktor Orban, then Hungary’s prime minister. According to reporting by the Financial Times, the SDA recognized that overt political or financial support originating from Moscow would immediately be interpreted as foreign meddling and severely damage Orban’s electoral prospects. Consequently, the firm proposed contacting local influencers through intermediaries, deliberately laundering Russian talking points so they appeared entirely homegrown.
The SDA has even bridged the gap between virtual online spaces and real-world physical actions by staging controversial or threatening events designed purely to generate media attention. In September 2025, severed pig heads—several of which were explicitly marked with the name of French President Emmanuel Macron—were left outside mosques and Islamic centers in and around Paris. Leaked SDA files later exposed the meticulous operational planning behind the stunt, alongside similar vandalism plots executed in other parts of France and Germany. Mainstream broadcast networks and social media platforms immediately covered the unsettling incidents, igniting widespread public anxiety regarding xenophobia and Islamophobia across Europe. The SDA internally refers to these calculated disruptions as "cognitive strikes," where the ultimate objective is securing maximum visibility, shaping public perceptions, and provoking intense emotional responses that can be politically exploited.
Such covert kinetic activity runs parallel to Moscow’s ongoing efforts to expand its official broadcast propaganda apparatus worldwide. In 2025, the Kremlin allocated 31 billion rubles—approximately $367 million—to RT, the Russian state-backed broadcaster. According to the BBC, RT and Sputnik, the state news agency and radio service, have aggressively expanded their footprints across Africa, Latin America, the Middle East, and Southeast Asia, frequently by absorbing or partnering with local media channels. The RT Academy, the broadcaster’s educational arm, even operated a journalism training program in the Indonesian capital of Jakarta to train a new generation of local reporters under the guise of teaching "fact-checking" and "ethical standards" in storytelling.
Beyond direct broadcasting, RT actively funds and operates sympathetic proxy channels. In 2024, the U.S. Department of Justice indicted two RT employees for funneling nearly $10 million to Tenet Media, a right-wing U.S. digital media company, though the prominent influencers working for the platform were reportedly unaware of the true source of their funding.
A CONFLUENCE OF INFLUENCE
Full-spectrum propaganda presents an extraordinarily distinct security challenge because it exploits the fundamental ways modern information ecosystems confer legitimacy. Psychological research has consistently demonstrated that sheer repetition reinforces belief; a fabricated narrative invariably appears far more believable when it surfaces independently across multiple disparate platforms, or when a seemingly unconnected network of individuals begins discussing it simultaneously. A false claim suddenly acquires an aura of credibility when an established outlet, a handful of prominent influencers, and anonymous voices within a Telegram channel all appear to arrive at the same conclusion without coordination.
Integrated propaganda campaigns are also uniquely adept at evading existing enforcement crackdowns enacted by individual tech platforms. Because social media corporations vary wildly in their regulatory policies and enforcement mechanisms, when fake accounts are purged from one site, they frequently respawn or migrate to alternative platforms with more permissive guidelines or weaker operational security. For instance, when Russian officials and state media networks portrayed the devastating bombing of a Ukrainian maternity hospital in March 2022 as a staged event staffed by paid actors, major platforms like Twitter swiftly removed the deceptive material. However, anonymous Telegram channels posing as objective fact-checking organizations quickly picked up the slack, gaining massive popularity by propagating the false narrative. While some of these channels may have been genuine independent investigations, many others were subsequently identified as the work of state-backed information warfare groups seeking fertile ground to distribute their messaging.
Consequently, analyzing foreign influence campaigns on a platform-by-platform basis—or treating them purely as a niche social media moderation problem—is profoundly inadequate. Scrutinizing posts on Bluesky, for instance, entirely overlooks how those specific messages connect to simultaneous, synchronized efforts on Facebook and X, articles published in state-controlled newspapers, and international public broadcasts. Comprehending a modern propaganda campaign requires a panoramic view: investigating where else a specific narrative has surfaced, whether fabricated news outlets were quoted by third-party influencers, whether ostensibly independent sources share hidden ties, and how these fragmented components reinforce one another.
The rapid maturation of artificial intelligence is further transforming these operational strategies, compounding the difficulty of detection and attribution. An extensive investigation by The New York Times revealed that China and Iran have utilized open-source Chinese AI models to execute automated propaganda campaigns through AI agents requiring minimal human supervision. These autonomous agents generated fake accounts across a diverse array of social media networks, published fabricated news stories, and coordinated messaging strategies in real-time. Although these early deployments had limited direct reach, they serve as dangerous precursors of future capabilities.
State media channels are similarly leveraging AI to scale their international reach. Researchers at Anthropic discovered that Russian state media utilized the Claude AI model to translate Romanian and Moldovan news articles into Russian, demonstrating how hostile states can effortlessly overcome linguistic barriers without requiring specialized human translators. Propagandists are even directly targeting AI recommendation and search systems to shape the outputs provided to everyday users. By creating fraudulent websites designed to surface when AI models retrieve web information, bad actors attempt to skew the foundational data used to train AI models on how to respond to sensitive queries.
CAPACITY COLLAPSE
Because U.S. adversaries are aggressively pursuing a full-spectrum, AI-optimized approach to information warfare, Washington urgently requires maximum institutional capacity to assess the threat holistically, expose covert networks, and counter them effectively. Tragically, however, both Silicon Valley technology firms and the U.S. federal government have spent the past several years dismantling those exact capabilities, leaving the nation largely blind to its underlying vulnerabilities.
Within Silicon Valley, major corporations have deliberately restricted access to platform data, severely limiting public visibility into ongoing propaganda operations. Social media companies have quietly rolled back platform integrity commitments, reduced transparency initiatives, and dissolved formal collaborative partnerships with independent researchers. In May, Meta executed a significant workforce reduction, laying off ten percent of its personnel—including members of its core investigations team, which had historically played an instrumental role in identifying and dismantling online influence networks. Furthermore, Meta retired CrowdTangle in 2024, a vital transparency tool that researchers relied upon to track public posts and measure the velocity of content. In 2023, Twitter (subsequently rebranded as X) eliminated free academic access to its application programming interface (API), which researchers routinely used to study public discourse. X now imposes steep commercial fees—charging $5 for every 1,000 tweets, amounting to a prohibitive minimum of $42,000 per month for enterprise subscriptions—making large-scale academic research financially impossible for most universities.
These self-imposed barriers have crippled independent verification of claims made by both tech platforms and government bodies regarding foreign influence, making it exponentially harder to gauge the true severity of hostile campaigns. For instance, in August, when X announced the sweeping removal of 200,000 accounts linked to a suspected Chinese bot farm—claiming that 200 of those accounts had posted content concerning domestic AI data centers—the company failed to release underlying operational data to researchers. While opportunistic politicians immediately seized upon the announcement to blame Chinese covert influence for genuine grassroots political backlash against domestic data centers, a subsequent analysis conducted by Clemson University professor Darren Linvill revealed that the identified posts had received virtually zero organic engagement and posed no meaningful influence.
Simultaneously in Washington, an intense political focus on combating alleged domestic censorship has triggered the wholesale dismantling of federal institutions dedicated to tracking foreign interference. President Donald Trump’s executive order targeting government censorship, signed on his first day back in office in January 2025, was publicly framed as an effort to safeguard constitutionally protected free speech. In practice, however, the policy has heavily suppressed federal initiatives aimed at mitigating hostile foreign influence, with broad allegations of censorship weaponized to defund or dissolve essential defensive structures.
In 2024, Congress allowed the official mandate of the State Department’s Global Engagement Center to expire. Established in 2016 with bipartisan support, the office’s primary mission was to track and expose state-sponsored propaganda originating from China, Iran, Russia, and other foreign adversaries. Although a replacement entity—the Counter Foreign Information Manipulation and Interference hub—was briefly created, it was subsequently shut down by the State Department. While conservative lawmakers and critics who campaigned against these offices argued they posed a threat to American civil liberties, no court of law ever found that the agencies had engaged in illegal censorship.
The contraction continued in August 2025, when Director of National Intelligence Tulsi Gabbard closed the Foreign Malign Influence Center, an organization that had coordinated the U.S. intelligence community’s cross-agency analysis of foreign interference methods. Concurrently, the Department of Justice shuttered the FBI’s Foreign Influence Task Force, which had been tasked with exposing unlawful foreign interference operations on U.S. soil while aligning investigations and intelligence-sharing protocols across federal agencies.
Combined, the systematic dismantling of these specialized institutions has created a profound intelligence vacuum in Washington regarding the true scope of operations being conducted by China, Iran, and Russia. As Senator Mark Warner, vice chairman of the Senate Select Committee on Intelligence, warned this spring, dangerously little institutional capacity remains to track or disrupt the sophisticated covert campaigns deployed by America’s geopolitical rivals. Just as foreign influence operations are becoming more seamlessly integrated across the digital and physical realms, the U.S. government is actively discarding the bureaucratic coordination required to comprehend them.
As the United States voluntarily blinds itself, foreign adversaries are rapidly accelerating their efforts. Future influence operations will increasingly rely on autonomous, AI-enhanced systems capable of instantly spinning up fake news outlets, generating synthetic articles and videos, disseminating them via hijacked accounts, seeding them among paid influencers, validating them through state-controlled broadcasters, and tricking AI search engines into treating falsehoods as objective truth. While China, Iran, and Russia operate fluidly across the entire spectrum of modern propaganda, the United States remains trapped in narrow, disconnected silos. Without the basic institutional capacity to understand the true nature of the threat, Washington cannot even begin to mount an effective defense.
