Its irritating that content creators, of all forms, are making money on engagement that isnt real. Fake consensus shaping narratives and perceptions is very effective especially on young people and im genuinley concerned. ~66% of engagment on hot button issues being artificial is a huge problem.

ChatGPT:

There isn’t one definitive number, because “artificial engagement” includes several different things:

• fully automated bots
• fake followers/likes/comments bought from click farms
• coordinated engagement groups (“pods”)
• AI-assisted accounts run partly by humans
• spam amplification systems
• inactive or zombie accounts inflating metrics

But most serious research lands in a broad range:

• Roughly 10–20% of social media accounts are estimated to be fake, bot-operated, or primarily inauthentic. (Nature)
• In influencer ecosystems specifically, studies often find 15–40% of engagement or followers showing suspicious or artificial patterns. (SociaVault)
• On highly politicized or viral topics, bots can dominate amplification. One Pew study estimated 66% of tweeted links to popular websites were shared by suspected automated accounts. (Pew Research Center)

So if you’re asking:

“What percentage of what I see online is artificially amplified?”

A realistic answer is:

• probably a meaningful minority overall
• and on some topics/platforms, possibly the majority of visible engagement

Especially in:

• politics
• crypto
• celebrity fandoms
• outrage content
• influencer marketing
• “viral” culture-war posts
• scam advertising

Why platforms permit it comes down to incentives.

1. Engagement = revenue

Social platforms make money from:

• ad impressions
• time-on-platform
• growth metrics
• investor perception

Artificial engagement still boosts:

• trending activity
• viewing time
• perceived popularity
• ad inventory

Even fake engagement can help make a platform look alive.

2. Detection is genuinely difficult

Modern bots no longer look like obvious spambots. Many are:

• partially human-operated
• AI-assisted
• coordinated through real accounts
• designed to imitate normal behavior

Researchers note that even humans are bad at detecting bots. (Sage Journals)

3. Platforms fear false positives

If a platform aggressively removes suspicious accounts:

• real users get banned accidentally
• creators get angry
• engagement numbers drop
• accusations of bias appear

So companies often tolerate a “background level” of fake activity.

4. Some artificial activity is actually useful to platforms

Not all bots are malicious:

• customer-service bots
• recommendation systems
• automated posting tools
• marketing automation
• news alerts

The line between “automation” and “manipulation” is blurry.

5. The industry became structurally dependent on inflated metrics

Influencer marketing, political campaigning, and even journalism often rely on:

• follower counts
• impressions
• virality metrics

There’s a huge economic ecosystem around manufactured social proof.

Some estimates put influencer fraud losses in the billions annually. (Fake Follower Checker)

The deeper issue is that social media ranking algorithms reward engagement signals, not truth or authenticity. If outrage, bot swarms, or fake comments trigger the algorithm, the system often amplifies them before moderation catches up.

That’s why many researchers now argue that modern social media is less a reflection of public opinion and more a competition between amplification systems — some human, some artificial.