The Illusion of Authenticity on X
From AI spam to state propaganda, X thrives on noise. Yet it’s still where leaders, celebrities, and entrepreneurs go viral first.
Every year since Elon Musk took over Twitter, one question keeps coming back: can we still trust what we see on X?
In 2025, the answer is layered. On the surface, X looks more open and dynamic than ever. Anyone can post. Anyone can go viral. But dig deeper, and the foundation is shaky. Bots, AI-driven accounts, extremist echo chambers, propaganda disguised as news, and harassment lists shape large parts of the platform.
Authenticity has collapsed. Relevance has not. X is still the global stage where leaders, celebrities, and entrepreneurs set the tone of public conversation.
1. The Jungle of X
We first launched the Authenticity Score two months ago on LinkedIn. The idea was simple: cut through the noise and measure whether someone’s influence was real or just manufactured.
After seeing the impact on LinkedIn, we decided to bring the same framework to X, but with a few important tweaks.
On LinkedIn, the main issue was pods, fake engagement groups, and AI-generated comments. On X, the environment is harsher. Here, propaganda impersonates official voices, extremist clusters create false consensus, and monetization farming floods the feed with low-effort spam.
That’s why we adapted the score for X, weighting identity, follower dynamics, and trustworthiness even more heavily. The result is a version built specifically for a platform where authenticity is collapsing, but reach is still unmatched.
Bot or human. Does the account reflect a real person or a fabricated identity?
Trustworthiness. Are claims backed by sources or built on emotional triggers?
Engagement. Is there organic debate or just echo chamber replies?
Follower dynamics. Do audiences grow naturally or through manipulation?
Content quality. Does content bring new insights or recycle templates?
The findings are stark: authentic behavior on X is the exception, not the rule.
When we break it down with the Authenticity Score, most accounts don’t fall into a gray area: they fall neatly into a few patterns of inauthenticity.
Here’s the full breakdown.
1.1 AI-Generated Ghosts
These are the most basic form of inauthentic accounts on X: profiles entirely run by AI, with not necessarily any human presence behind them. Their sole purpose is to post continuously, flooding the feed with content in the hope that repetition alone will generate reach.
Take Succeeded Mind as an example. On the surface, it looks like a classic self-help or business coaching account. In reality, it posts 24/7 with generic motivational quotes and business clichés. Every tweet follows the same recycled template: “5 lessons to grow faster,” “10 mistakes to avoid,” “The secret to success is…”
The signals are unmistakable:
No human presence. There are no photos, no personal anecdotes, no “this happened to me” stories. Nothing that anchors the account to a real life.
Always on. Posts appear every hour of the day, every day of the week. A real human simply couldn’t sustain that rhythm.
Recycled ideas. There’s no novelty. The account cycles endlessly through generic business or self-improvement tropes, slightly rephrased to look new.
Why they matter: they flood the platform with low-value noise. X’s algorithm doesn’t care whether a post is insightful or not — it only sees engagement and volume. AI-ghost accounts exploit that, overwhelming timelines and burying authentic voices under layers of synthetic filler.
This is the most obvious form of inauthenticity: automation pretending to be inspiration.
1.2 Engagement Farmers
If AI ghosts are synthetic, engagement farmers are the human equivalent: real people, but posting with zero authenticity.
These users have figured out that X’s monetization program pays for impressions and replies. The formula is simple: the more eyeballs and engagement, the more money. So instead of sharing meaningful ideas or lived experiences, they farm engagement with the lowest-effort content possible.
The tactics are predictable:
Bait questions. “Tea or coffee?” “Walk or crawl?” “Which is better: sunrise or sunset?” Questions so simple that anyone can reply without thinking.
One-liners. Provocative statements that trigger immediate emotional responses: not because they are deep, but because they are designed to polarize.
Filler threads. Generic advice threads with flashy hooks but no substance, repackaged from AI or copied from other creators.
Why they matter: engagement farming is spam for profit. Unlike AI ghosts, these accounts are human-run, but their intent is no less inauthentic. They game the payout system, chasing impressions and replies as a financial strategy.
The result: inflated metrics. A post that generates a million impressions might look successful, but the replies are shallow, the conversation nonexistent, and the influence fabricated. As one creator noted in April 2025, payouts on X have been “ruined” by operators farming engagement at scale, particularly from India.
These accounts are not fake in identity. But they are fake in purpose. They treat X not as a place to exchange ideas, but as a machine to be exploited.
1.3 Extremist Amplifiers
The third type is more dangerous: the Extremist Amplifiers.
These accounts are run by real people, often with strong ideological positions. But their influence doesn’t come from authentic debate. It comes from being embedded in tight extremist clusters.
The pattern is easy to trace:
An account posts a provocative take, often inflammatory, often tied to identity, nationalism, or cultural conflict.
Within minutes, a wave of replies floods in. Always the same accounts, always aligned with the same ideology.
The replies don’t challenge, question, or expand. They repeat short, prepackaged phrases: “So true,” “Wake up,” “Disgusting.”
What looks like broad consensus is really a small circle amplifying itself. The repetition creates the illusion of mass support, and the algorithm interprets that as engagement worth spreading.
Why they matter: repetition normalizes. Fringe views, echoed loudly and consistently enough, begin to look mainstream. This is not accidental, it’s a strategy. Extremist amplifiers exploit the mechanics of virality to make radical positions feel common, slowly reshaping the perception of what’s normal.
Unlike AI ghosts or engagement farmers, extremist amplifiers don’t just fill space. They shift discourse.
And yet, these accounts are not marginal. RadioGenoa ranks in the Top 50 most influential accounts on X worldwide. Its Authority Score is over 9,000, the same range as global celebrities and political leaders.
That is the brutal reality:
You don’t need authenticity to be powerful on X.
Virality beats trust.
Emotional triggers beat originality.
Influence is no longer tied to credibility.
1.4 State & Political Operators
Finally, the most sophisticated group: the State and Political Operators.
These accounts are not run by random individuals or profit-seekers. They are tied — directly or indirectly — to coordinated propaganda campaigns. Their purpose is not visibility or payouts. Their purpose is influence.
Some impersonate governments, like the “Russian Army” account, using military insignia and national flags to appear legitimate. Others, like RadioGenoa, frame themselves as regional commentators or grassroots voices, while in reality pushing orchestrated narratives.
The tactics are consistent:
Borrow credibility. Use symbols of authority — flags, institutional handles, “official”-sounding names.
Exploit virality. Post emotional and dramatic predictions designed to spread fast — “Ukraine will collapse tonight,” “The final strike is coming.”
Scale through coordination. These accounts don’t go viral randomly. They rely on cross-border clusters, waves of amplification, and echo networks that push their content far beyond what a normal account could achieve.
Why they matter: they are not just noise. They are weapons of influence. These accounts play a direct role in shaping global perception, bending the narrative on conflicts, elections, and political movements. Unlike engagement farmers, they are not just gaming the system — they are weaponizing it.
A pro-Russian disinformation campaign dubbed Operation Overload (or Matryoshka) ramped up AI-generated content targeting Ukraine and Western audiences. According to recent findings, between September 2024 and May 2025, the volume of misleading content ballooned from 230 to 587 unique items, including deepfake videos, AI-generated images, and fabricated narratives shared via bot networks on X. (Source: Washington Post)
But not all coordinated influence looks like anonymous bots or fake government accounts. Sometimes it wears the mask of a real human presence.
This is where Olia comes in.
She presents herself as a Jewish advocate and political commentator. Unlike Succeeded Mind or Russian Army, this account does appear to be operated by a real human, the voice is consistent, and posts mix personal identity markers with political commentary.
But the authenticity stops there. The content patterns raise red flags:
Trustworthiness (15/100): Posts present strong claims about ethnic and religious issues but rarely provide evidence or verification. Emotional framing dominates.
Engagement & Replies (25/100): Replies cluster within minutes, often showing the same supportive accounts repeating predictable responses. This suggests coordination rather than organic debate.
Follower Dynamics (35/100): Audience growth is suspicious. The account attracts clusters of followers aligned around Israel–Middle East messaging, pointing to external amplification.
Content Quality (20/100): Posts rely on inflammatory framing, templated outrage, and repetitive use of talking points. Very little original insight, personal experience, or constructive commentary is shared.
Olia demonstrates that not all inauthentic activity is anonymous or bot-driven. Sometimes, real humans act as amplifiers within coordinated influence campaigns.
Unlike Succeeded Mind, this is not about AI spam.
Unlike Russian Army, this is not impersonation.
Unlike RadioGenoa, this is not propaganda dressed as a top influencer.
Instead, this is a human voice shaped by external coordination. The presence of real identity markers gives it credibility, but the messaging patterns — repetitive, inflammatory, amplified by the same clusters — suggest it is part of a larger organized influence ecosystem.
This use case shows how blurry the line between “authentic person” and “coordinated operator” can be. An account can be run by a real individual, but function as a node inside a wider networked campaign.
This is where the line between “platform dysfunction” and “geopolitical manipulation” blurs. The jungle stops being annoying, and starts being dangerous
1.5 Harassment Lists
On top of these categories, X has another layer of dysfunction: lists.
Users create lists with names like “Rape apologists,” “Far Left Mongaloids,” or “Engagement farmers.” Unlike hashtags or groups, lists are not neutral. They don’t categorize for discovery. They stigmatize.
Being added to a list makes you a target. Other users pile on with harassment.
Discovery gets skewed. Instead of surfacing relevant voices, lists amplify abuse.
Abuse becomes normalized. When slurs sit in list titles, they become part of the everyday visual language of X.
Why they matter: Lists weaponize visibility. They don’t help users filter useful voices. They mark them for attack.
2. The Nature of the Jungle
Put together, these categories explain why X feels chaotic yet strangely patterned. On the surface it looks like noise — random posts, random fights, random virality. But once you map the behavior, the pattern is clear. It is structured noise.
AI ghosts flood the platform with volume. They ensure there is always something to scroll, even if it has no substance. The algorithm sees activity, not quality, so these accounts keep circulating.
Engagement farmers exploit monetization. They are not bots but humans who figured out the payout loopholes. Their content is engineered not to inform, but to bait: endless low-effort questions, viral one-liners, or polarizing prompts. Their goal is not conversation but profit farming, inflating engagement metrics to trigger payouts.
Extremist amplifiers reinforce each other’s narratives. They make fringe opinions look mainstream. By replying in unison, they simulate consensus. The emotional weight of repetition convinces outsiders that “everyone is talking about this,” when in reality it’s a closed loop.
State operators coordinate propaganda. These accounts weaponize the structure of X. They mimic authority, borrow legitimacy through flags and titles, and repeat dramatic claims until they feel credible. Their power is not in being believed at once, but in wearing down skepticism through volume.
Lists channel harassment. They weaponize discovery. Once you’re added to a hostile list, you become a target. Lists don’t create debate, they create mobs. They stigmatize voices and shrink the space for normal users who don’t want to risk being targeted.
On top of this structure is the policy environment. Under Musk, X has embraced “freedom of speech” as its core value. In practice, that has meant fewer filters and almost no friction for spam, harassment, or propaganda. Bad actors face little resistance because the platform is reluctant to moderate. The result is not more open debate, but more space for inauthentic behavior to thrive.
The outcome: X rewards scale and emotion over trust and truth.
Volume wins over originality.
Anger wins over nuance.
Repetition wins over verification.
This is why the platform feels like a jungle. It’s not because it is uncontrolled, but because it is optimized for the wrong incentives. Payout systems reward spam. Algorithms reward outrage. And “freedom of speech” in practice removes the last filter that could slow the spread of inauthenticity.
The jungle is not an accident. It is the system working as designed.
4. Counterpoint: When Authenticity Thrives
The jungle is real, but not universal. Some niches succeed precisely because they are authentic by definition.
Entrepreneurs & Build-in-Public
Unlike propaganda clusters or AI spam accounts, entrepreneurs building in public use their real stories as content.
Take Rob Hallam:
Goal: build a $10K/month SaaS while traveling the world.
Posts transparently about risks (“Left my job. Left my home. Living on savings.”).
Shares milestones, failures, pivots, and personal reflections.
Viral reach: his pinned video has nearly 1 million views.
For these creators, authenticity is not optional: it is the product. The appeal comes from following a real founder in real time, watching wins and mistakes unfold.
Favikon data supports this:
B2B creators on X grow 3x faster than on LinkedIn.
Entrepreneurial content, especially when transparent, drives outsized engagement compared to AI spam or templated posts.
This niche proves that authentic content can still cut through the jungle, because it is impossible to fake the personal risk and transparency of build-in-public
5. Why X Still Matters
Despite bots, spam, and propaganda, X remains the main stage for global influence.
Look at the Favikon Top 200 creators on X:
Justin Bieber — 108.9M followers
Elon Musk — 224.6M followers
Cristiano Ronaldo — 115.7M followers
Barack Obama — 130.5M followers
Narendra Modi — 98.0M followers
BTS — 45.4M followers
Taylor Swift — 93.7M followers
These are not marginal voices. They are the cultural, political, and economic leaders of the world.
And their stage is X.
86 of the Top 200 are from the United States, showing the platform’s central role in shaping not just global influence, but American dominance in digital culture.
The AI Example
If there’s one niche that shows the dominance of X, it’s artificial intelligence.
Look at the Favikon Top 200 in AI Research & Innovation:
Lex Fridman (MIT researcher & podcast host) → 4.8M followers on X, the majority of his audience sits there.
Sam Altman (CEO of OpenAI) → 3.9M followers on X, where he makes announcements, debates AI policy, and tests ideas in real time.
Aravind Srinivas (CEO of Perplexity AI) → 292K on X, compared to 58K on LinkedIn. Growth on X is explosive (+12.88%).
Andrej Karpathy (ex-Tesla, ex-OpenAI) → 1.4M on X, vs just 84.5K on LinkedIn.
Andrew Ng (Coursera, Stanford) → 2.2M on X, dwarfing his 1.2M LinkedIn following.
Yann LeCun (Meta & Turing Award winner) → 944K on X, compared to just 84.6K on LinkedIn.
The pattern is obvious: AI leaders choose X as their main platform.
Even though LinkedIn is bigger for professional networking, when it comes to thought leadership and influence, the AI conversation happens on X. Why?
Real-time debate. Policy announcements, research papers, and product launches trigger immediate responses.
Direct connection. CEOs, researchers, engineers, and investors engage each other directly without intermediaries.
Virality. Technical content can reach millions overnight, something LinkedIn’s slower, algorithm-curated system cannot replicate.
This shows why X is still indispensable: even in a professional niche where LinkedIn should dominate, the most influential voices overwhelmingly use X. It’s not just entertainment or politics, X remains the default stage for global innovation conversations.
For all its flaws, X is still where announcements break first, where movements spark, and where entrepreneurs can test ideas in real time.
Why the Competition Failed
Over the past two years, multiple challengers tried to dethrone X. All failed to build lasting momentum.
Threads (Meta). Launched with fanfare in 2023, it had the reach of Instagram but none of the urgency of X. Conversations felt slow, discovery was weak, and newsmakers didn’t migrate. Engagement collapsed after the initial hype.
Bluesky. Marketed as a decentralized alternative, it attracted early adopters and tech enthusiasts but never broke into the mainstream. Without celebrities, politicians, or entrepreneurs posting daily, it became a niche community rather than a global platform.
Mastodon. Open-source and decentralized, it appealed to those who wanted an ethical alternative. But the fragmented server model confused new users, and it never reached the scale required to be relevant in global conversations.
All three faced the same problem: they couldn’t replicate the immediacy of X.
When news breaks, when a political leader makes a statement, when an entrepreneur announces a product, it happens on X first. Journalists, investors, policymakers, and celebrities are already there, refreshing in real time.
That density of attention is impossible to copy. Competitors may build friendlier platforms, but they can’t recreate the network effect of having the world’s most powerful people in one place, posting directly and unfiltered.
6. Verdict
So, is X authentic in 2025
No.
AI shells flood the feed with recycled content.
Extremist echo chambers create false consensus.
State-backed propaganda impersonates officials and spreads disinformation.
Even a Top 50 account like RadioGenoa scores 14/100 on authenticity.
But is X still relevant?
Absolutely.
It remains the fastest way to reach millions, the stage for world leaders, and the platform where entrepreneurs can build momentum overnight.
The lesson is simple: X in 2025 is not about trust. It is about reach.
And in influence, reach is still everything.
That said, I don’t believe X can continue like this forever. Running a global information platform with zero meaningful moderation is unsustainable. The jungle works in the short term because it maximizes virality, but at some point, the chaos undermines the credibility of the entire system.
We already see it happening. B2B creators have left X. Many of the voices that made this space vibrant in 2020–2022 have shifted to LinkedIn or Substack. The marketing, business and HR niches, once thriving on Twitter, have virtually vanished from X. They want stability, a safer environment, and algorithms that reward expertise rather than spam. For B2B, it’s a massive loss, and LinkedIn is gaining momentum from it.
If X doesn’t find a balance between “freedom of speech” and basic authenticity standards, it risks breaking under its own contradictions. Right now, reach keeps it relevant. But without trust, reach can only carry it so far.
Wow, this is a huge research. I like X for insights from entrepreneurs and as a fastest way to read news.