Contacts
1207 Delaware Avenue, Suite 1228 Wilmington, DE 19806
Contact
Business Address: 1207 Delaware Avenue, Suite 1228 Wilmington, DE 19806

AI in Social Media Statistics 2026: 94% of Social Teams Use AI, 3 Billion Labeled TikToks and a 91-Point Disclosure Gap

AI in social media statistics 2026, SAPI gauge at 46.1 with the four layer bars. Alt: "Social AI Penetration Index 2026 reads 46.1: creation 94, audience 56, ad buying 31.6, disclosure 2.7, 94% create with AI · 2.7% labeled" split visual. Alt: "94% of social media marketers use AI while about 2.7% of daily TikTok uploads carry an AI label

AI in Social Media Statistics 2026

By Axis Intelligence Research

Co-authors: Elena Rodriguez & Alex Rivera | Last updated: October 20, 2026 | License: CC BY 4.

Axis Intelligence Research estimates that AI-automated campaigns now account for at least 31.6% of Meta’s advertising run rate, while TikTok’s own disclosures imply about 7.3 million AI labels applied per day. Yet only about 2.7% of daily TikTok uploads carry a label, against 94% of social marketers using AI in their workflow.


Quick Answer

In 2026, 94% of social media marketers use AI in their workflow (HubSpot), up from 71% a year earlier. Meta says its Advantage+ AI ad products run at more than $75 billion a year, which Axis Intelligence Research calculates as at least 31.6% of Meta’s annualized ad revenue. TikTok has labeled over 3 billion videos as AI-generated, and 56% of users in the US, UK and Australia say they see “AI slop” often or very often (Sprout Social). Our composite measure, the Social AI Penetration Index (SAPI), reads 46.1 out of 100.

Key Findings

  1. Axis Intelligence Research calculates that Meta’s AI-automated Advantage+ ad products represented at least 31.6% of Meta’s annualized advertising revenue in Q2 2026, up from 30.0% in Q3 2025.
  2. TikTok went from over 1.3 billion to over 3 billion AI-labeled videos in 233 days, which Axis Intelligence Research calculates as roughly 7.3 million new AI labels per day.
  3. While 94% of social media marketers use AI, only 13.54% apply it to social listening and 9.75% to brand voice auditing, according to HubSpot’s 2026 data.
  4. Half of Gen Z social media users (50%) have unfollowed, muted or blocked an account for posting AI slop, compared with 29% of Baby Boomers (Sprout Social, n=2,250).
  5. The Social AI Penetration Index (SAPI) reads 46.1 out of 100 as of October 20, 2026, with a 91.3-point spread between AI use in content creation (94) and labeled AI supply on TikTok (2.7).

How Many Social Media Marketers Use AI in 2026?

Adoption is close to saturation. HubSpot’s 2026 Social Media Trends Report, a survey of 1,007 social media marketers in North America, Europe, Asia and Australia fielded in February 2026, puts AI use in social workflows at 94%, against 71% in 2025 and 73% in 2024 (HubSpot). After a flat 2025, adoption jumped 23 points in a single year. A smaller agency-heavy survey by Sociality.io lands on a similar pattern: 64.1% use AI daily and 89.7% at least weekly (Sociality.io).

That second figure is widely quoted, and it deserves a caveat we explain in the evidence audit below: the respondent base is 39 people.

MetricValueAs ofSource
Social marketers using AI in workflows94%Feb 2026HubSpot
Same metric, 2025 survey71%2025HubSpot
Same metric, 2024 survey73%2024HubSpot
Use AI daily64.1%Jan 2026Sociality.io
Use AI at least weekly89.7%Jan 2026Sociality.io
Never use AI2.6%Jan 2026Sociality.io
Expect AI budget to increase in 202661.5%Jan 2026Sociality.io

What do social media teams use AI for?

Most AI use sits at the top of the content funnel: ideas and words. In 2025, AI-generated short-form video (56%) and image generation (53%) were the top uses; in 2026, HubSpot finds 42.9% use AI for ideation and 41.5% for captions and short-form text. Image generation (38.35%) and image editing (34.78%) follow. Sociality.io’s respondents rank analytics and reporting (59.5%) level with ideation (59.5%), ahead of captions (45.9%) and visual or video creation (40.5%). Workflow automation is marginal at 5.4%.

AI use case (HubSpot, 2026)Share of social marketers
Ideation and brainstorming42.9%
Captions and short-form text41.5%
Image generation38.35%
Image editing34.78%
Social listening and trend monitoring13.54%
Brand voice auditing9.75%

Source: HubSpot 2026 Social Media Marketing Report and HubSpot social media trends.

Where does AI fall short for social media teams?

The tasks marketers adopt most eagerly in visuals are also the ones that disappoint most. HubSpot reports that AI fell short of expectations for 40.2% of those using it for image generation, 29.58% for image editing and 28.7% for video generation. Sociality.io’s respondents mirror this: 78.4% say AI output needs moderate or extensive editing before publishing, and only 28.2% say more than half of their posts are AI-assisted. Only 19.4% list disclosure and transparency as a concern, well behind brand voice consistency (30.6%).

The worry is less about speed than about voice. 77% of HubSpot respondents rank authenticity above production value, and 16% name “implementing AI without sacrificing quality or trust” as their single biggest challenge. In Sociality.io’s sample, originality (61.1%) and accuracy (50%) are the top concerns, while time savings (71.1%) is the top benefit.

Which AI tools do social media marketers use?

General-purpose chatbots dominate. HubSpot finds 71.07% of social marketers use AI chatbots; by tool, 68.46% use ChatGPT, 47.43% Google Gemini and 17.63% Claude. Sociality.io’s ranking is consistent: conversational chatbots 69.2%, visual tools 59%, text generators 41%.

Are social media marketers abandoning AI image and agent tools?

Standalone tools are losing ground to chatbots. HubSpot’s full report shows that in one year, use of visual AI tools such as Midjourney and Adobe Firefly fell from 53% to 31%, and use of AI agents and assistants fell from 50% to 32%, while chatbot use rose from 52% to 71%. Video AI tools (27%) and AI built into the social platforms themselves (24%) trail behind.

AI tool category (HubSpot)20252026
Chatbots52%71%
Visual AI53%31%
AI agents and assistants50%32%
Video AINA27%
In-platform AI toolsNA24%

Source: HubSpot 2026 Social Media Trends Report (n=1,007, February 2026). The 2025 values are restated in the 2026 report.

For a broader marketing view beyond social, see our AI marketing statistics and content marketing statistics.

Analyst view — Sarah Mitchell. The 94% figure tells us AI is in the building, not that it runs the building. The drop in visual AI and agent tools is the overlooked signal: marketers tried dedicated generators and gave up on many of them, in line with the 40.2% who say image generation fell short. The adoption curve is lopsided: text and ideation are routine, while the two uses that most protect a brand on social, listening (13.54%) and voice auditing (9.75%), are barely touched. That is the most actionable gap in this dataset. Teams generating more posts with AI but not using AI to check what audiences are saying back are scaling output without scaling feedback, which is exactly the condition under which “AI slop” complaints grow.

How Much of Meta’s Ad Business Runs on AI Automation?

Platform-side automation is where AI in social media is largest in dollar terms. On its Q2 2026 earnings call (July 29, 2026), Meta said its Advantage+ end-to-end AI ad solutions had passed a $75 billion annual run rate. Meta’s Form 10-Q for the quarter ended June 30, 2026 reports advertising revenue of $59.363 billion for the quarter.

Axis Intelligence Research calculation:

  • Annualized Q2 2026 ad revenue = $59,363M × 4 = $237,452M
  • AI-automated share = $75,000M ÷ $237,452M = 31.6% (a floor, because Meta says “over” $75 billion)
  • Same method for Q3 2025: $60,000M ÷ ($50,082M × 4) = 30.0%

Over those three quarters, Meta’s AI ad run rate grew 25.0% while its quarterly ad revenue grew 18.5%. AI-automated buying is taking share inside Meta’s own ad book, not just growing with it.

MetricQ3 2025Q2 2026Source
Quarterly ad revenue$50.082B$59.363BMeta IR, 10-Q
AI ad solutions annual run rate (company-reported)>$60B>$75BMeta IR
AI-automated share of annualized ad revenue30.0%31.6%Axis Intelligence Research calculation

How many advertisers use Meta’s AI tools?

Meta reported on the same call that more than 9 million small businesses used at least one of its AI ad creative tools, and more than 1 million businesses used Meta Business Agents every week to talk to customers or complete sales, after the agents became available globally on WhatsApp and Messenger. Meta also said its user-understanding models, combined with its GEM ad-ranking model, produced an 8.3% increase in Facebook ad clicks and a 15.7% uplift in conversions. These are company-reported figures and have not been independently verified. For scale, the 10-Q shows Q2 2026 ad revenue up from $46.563 billion a year earlier.

For Meta’s wider ad economics, see Facebook ads statistics, Instagram ads statistics and our Meta statistics hub.

Analyst view — Sarah Mitchell. Most “AI in social media” coverage measures what marketers type into chatbots. The money tells a different story: nearly a third of Meta’s ad business is now bought through a system where the platform’s AI chooses audiences, placements and often creative. For media buyers, the strategic question has shifted from “should we use AI?” to “what inputs do we still control?” Creative volume, first-party conversion signals and exclusion lists are the levers left.

How Many AI-Generated Videos Are on TikTok?

TikTok is the only major platform that publishes a running count of AI labels. On November 19, 2025, it reported having labeled over 1.3 billion videos as AI-generated content (TikTok newsroom). On July 10, 2026, the figure was over 3 billion, applied through C2PA Content Credentials, creator self-labels and invisible watermarking (TikTok newsroom).

Axis Intelligence Research calculation:

  • Labels added = 3.0B − 1.3B = 1.7B over 233 days
  • ≈ 7.3 million AI labels per day (both inputs are floors)

What share of TikTok uploads is labeled as AI?

TikTok does not publish daily upload volume. The most recent independent census, by researchers from McGill University and partners (Steel et al., arXiv 2504.13279), counted about 269.3 million posts on a single day in April 2024, of which at least 0.5% carried TikTok’s AI-generated tag.

Axis Intelligence Research estimate: 7.3M ÷ 269.3M ≈ 2.7% of daily uploads now receive an AI label, roughly 5.4 times the 2024 tagged rate. This estimate assumes 2024 posting volume; if uploads have grown, the true share is lower.

MetricValueAs ofSource
Cumulative AI-labeled videos>1.3BNov 19, 2025TikTok
Cumulative AI-labeled videos>3BJul 10, 2026TikTok
New AI labels per day≈7.3MJul 2026Axis Intelligence Research calculation
Posts per day (independent census)269.3MApr 2024Steel et al. [older data]
AI-tagged share of daily posts0.5%Apr 2024Steel et al. [older data]
Labeled share of daily posts≈2.7%Jul 2026Axis Intelligence Research estimate
Fake accounts removed>86MQ1 2026TikTok

TikTok also committed more than $4 million to AI literacy programs, reported over 200 million views of its AI literacy content since November 2025, and joined the C2PA Steering Committee.

How much of a new TikTok user’s feed is AI-generated?

Labeled supply and what a user actually sees are different measurements. In a test of a fresh account’s For You feed in May 2026, Kapwing classified 294 of 500 videos (59%) as AI slop, against 21% in its earlier YouTube Shorts test (Kapwing). Across 10,742 manually reviewed videos in 20 categories, the Kids category reached 57.4% and the #cartoonkids tag 97%, while Fashion (1.3%), Music (1.5%) and Fitness (1.6%) were nearly free of it. Science and Education (35%), Health (33.8%) and History (33.5%) sat in between.

Two cautions. Kapwing sells AI video tools, and the feed test used a single new account, whose recommendations are not representative of established users. Its 59% and our 2.7% are not contradictory: one measures a recommendation-weighted feed for a cold-start account, the other measures labeled uploads across the platform. The two figures sit an order of magnitude apart, and that distance is itself the story: most AI content a viewer encounters carries no label.

More on the platform in our TikTok statistics.

Analyst view — Sarah Mitchell. The categories where unlabeled AI content concentrates (kids, health, science, history) are the ones regulators already treat as sensitive. That is where the next wave of labeling enforcement is most likely to land, and brands publishing in those verticals should assume their AI-assisted posts will face higher scrutiny than lifestyle content. Our social media regulation statistics track that pipeline.

How Do Audiences React to AI Content on Social Media?

Audiences notice, and a meaningful share act on it. Sprout Social’s Q1 2026 Pulse survey (n=2,250 across the US, UK and Australia, fielded February 5–9, 2026) found 56% of users see AI slop often or very often, and 83% at least sometimes (Sprout Social survey analysis).

How often users see AI slopShare
Very often26%
Often30%
Sometimes26%
Rarely8%
Never9%

Source: Sprout Social Q1 2026 Pulse.

Do people unfollow accounts that post AI content?

Yes, and the reaction scales with age in reverse. 50% of Gen Z respondents have unfollowed, muted or blocked an account for posting AI slop, versus 44% of Millennials, 38% of Gen X and 29% of Baby Boomers. 66% say they are more selective about who they follow than a year ago, and the most-cited thing users want brands to stop doing is posting unlabeled AI content (28%). See Gen Z social media statistics for the generation’s wider habits.

Is AI eroding trust in news on social media?

88% of Sprout Social’s respondents agree AI video tools have reduced their trust in news on social media; 35% say their trust in social media fell over the past year, with AI slop cited by 20% as the reason. The Reuters Institute Digital News Report 2026 (48 markets, 97,520 respondents) shows the wider backdrop: 54% use social and video networks for news weekly, but only 22% trust news they find there, versus 37% for news overall, a 15-point gap (Axis calculation). Trust in AI chatbot news answers sits at 20%, while weekly use of chatbots for news rose to 10% from 7%. Concern about fake news online reached 62%.

In the US, Pew Research Center found in 2025 that 76% of adults say it is important to tell AI-made from human-made content, while 53% are not confident they can [older data].

Analyst view — Sarah Mitchell. The audience data reverses an assumption in many content calendars: that AI’s cost is only quality. Among younger users, it is reach. A 50% Gen Z unfollow rate for AI slop means undisclosed, low-effort AI output carries a measurable follower cost, and disclosure is what users are explicitly asking for. Labeling is no longer just a compliance task; it is a retention tactic.

The Social AI Penetration Index (SAPI): How Deep Is AI in the Social Media Stack?

No single survey describes how far AI has moved into social media, because adoption happens in four different places: what teams create, how ads are bought, what platforms disclose, and what audiences experience. The Social AI Penetration Index (SAPI), developed by Axis Intelligence Research, puts those four layers on one 0–100 scale.

Formula: SAPI = 0.25 × Creation + 0.25 × Ad buying + 0.25 × Disclosure + 0.25 × Audience

SAPI layerWhat it measuresInputValueWeightSource
CreationShare of social marketers using AI in workflowsSurvey94.025%HubSpot
Ad buyingAI-automated share of Meta annualized ad revenueAxis calculation31.625%Meta 10-Q
DisclosureLabeled AI share of daily TikTok uploadsAxis estimate2.725%TikTok, Steel et al.
AudienceUsers seeing AI slop often or very oftenSurvey56.025%Sprout Social
SAPI46.1Axis Intelligence Research

Calculation: (94.0 + 31.6 + 2.7 + 56.0) ÷ 4 = 46.1 (baseline reading, October 20, 2026).

What the SAPI reading means

A score of 46.1 says AI has moved past the experiment stage in social media but is far from uniform. The defining feature is the 91.3-point spread between creation (94) and disclosure (2.7). AI is nearly universal in how content is made, and nearly invisible in how it is labeled. The audience layer (56) sits between them: people perceive far more AI content than platforms tell them about.

Equal weights are deliberate. Each layer represents a different stakeholder (marketers, advertisers, platforms, users), and no public data justifies ranking one above another. Each future reading will use the same four inputs, updated as new releases appear, so movement in the index reflects real change rather than a change of method.

Analyst view — Sarah Mitchell. Watch the disclosure layer. It is the one most likely to move sharply, because EU AI Act transparency obligations and C2PA adoption both push platforms to label more. If disclosure climbs toward the audience layer, the trust gap narrows. If creation keeps rising while disclosure stays in single digits, expect the audience layer, and the unfollow numbers, to rise with it.

How Reliable Are the Most-Cited AI in Social Media Statistics?

Many “AI in social media” figures circulating online share three problems: no sample size, no date, or no traceable origin. We audited the sources behind this page and several widely repeated numbers.

SourceSample or baseTypeDateReliability notes
Reuters Institute DNR 202697,520 respondents, 48 marketsAcademic surveyJun 2026Strong; YouGov online panel
Pew Research Center5,023 US adultsProbability panelJun 2025Strong; older data
Sprout Social Q1 2026 Pulse2,250 users, 3 countriesVendor survey (Glimpse)Feb 2026Good; methodology published
HubSpot 2026 report1,007 marketers, 4 regionsVendor survey (Datalily)Feb 2026Good; methodology in gated PDF
Sociality.io 2026 reportNot published (39 inferred)Vendor surveyJan 2026Directional only
Kapwing TikTok report500-video feed test; 10,742 videosManual reviewMay 2026Single account; vendor sells AI tools
Steel et al.Full-day platform sampleAcademic censusApr 2024Strong method; older data
Ricker et al.14,989,385 profile imagesAcademic detector studyMar 2023Strong method; older data
Meta and TikTok disclosuresCompany-wideCompany-reported2025–2026Not independently audited

Is the “89.7% of marketers use AI weekly” statistic reliable?

It is accurate to its source, but the source is small. Sociality.io does not publish its sample size, including in the full PDF report. Every percentage on its frequency question resolves exactly to a denominator of 39: 64.1% = 25/39, 25.6% = 10/39, 7.7% = 3/39, 2.6% = 1/39 and 89.7% = 35/39. The respondent profile resolves to 39 as well (43.6% social media managers = 17/39; 46.2% at companies of 11–50 employees = 18/39), and later questions to 34–38 answers as some respondents skip (44.7% = 17/38; 78.4% = 29/37). This is an Axis inference, not a disclosed figure, but it holds on every question we checked.

The report also contradicts itself: 2.6% say they never use AI, yet 10.3% select “no AI tools” on the tool question. With 39 respondents, 59% of them at agencies and 46.2% at companies of 11–50 employees, the 89.7% figure is directional. HubSpot’s 94% (1,007 respondents) is the stronger adoption benchmark.

Is it true that most social media images are AI-generated?

We found no primary source for that claim. A figure stating that 71% of social media images are AI-generated circulates widely, yet we could not trace it to any study, dataset or company disclosure. The largest measured count we found points the other way for profile photos: Ricker et al. analyzed nearly 15 million Twitter profile images in March 2023 and identified 0.052% as AI-generated faces [older data]. Generation has accelerated since, and Kapwing’s 2026 feed test shows high concentrations in specific categories, but no published evidence supports a platform-wide majority for images.

Is AI Changing the Business of Social Media Management Tools?

Social media management software vendors are repositioning around AI agents, and their financials show where the growth sits. Sprout Social reported Q2 2026 revenue of $123.8 million, up 11% year over year (Sprout Social Q2 2026 results). Customers contributing $30,000+ in annual recurring revenue rose to 3,926 (+11%), and those above $50,000 to 2,127 (+16%). The $30,000+ cohort accounted for about 61.4% of subscription revenue.

The company launched its Trellis AI agent and a paid Trellis Plus tier in July 2026. Growth concentrated in large accounts suggests that AI-heavy features (listening, analytics, agents) sell best to enterprise teams, which is consistent with HubSpot’s finding that listening remains underused across the broader market. Growth in the larger cohorts outpaced overall revenue growth.

For the broader context, see our social media statistics, creator economy statistics and influencer marketing statistics.

Analyst view — Sarah Mitchell. The tool market is splitting. Content generation is commoditized by free chatbots; ChatGPT alone reaches 68.46% of social marketers. What vendors can still charge for is the layer chatbots do not do well: listening, cross-channel analytics and governance. Buyers evaluating social tools in 2026 should weight those capabilities over built-in caption generators, which their teams already have elsewhere.

Methodology

Axis Intelligence Research compiled this dataset in October 2026 using only primary sources: company filings and investor materials (Meta, Sprout Social), platform newsroom disclosures (TikTok), academic research (McGill-led census, Ricker et al.), an independent institute (Reuters Institute) and a nonpartisan research center (Pew), plus vendor surveys (HubSpot, Sprout Social, Sociality.io, Kapwing). Every URL was opened and checked on October 20, 2026. HubSpot’s full report, Meta’s Q2 2026 call transcript and earnings presentation, and Sociality.io’s full PDF report were checked line by line against the figures used here.

Each figure in the CSV records its as-of date, source organization, document, URL and retrieval date. Values computed by Axis are flagged axis_calculated = yes, with the formula in method_note. Figures older than 12 months are marked [older data] and kept only where no newer primary source exists. Company-reported figures (Meta run rates, TikTok label counts) are used as stated; where a company gives a floor (“over $75 billion”, “over 3 billion”), our derived figures are floors too.

The TikTok labeled-share estimate combines two sources with different dates and should be read as an order-of-magnitude indicator. The SAPI uses four equally weighted layers, each already expressed on a 0–100 scale, with no further normalization.

About This Dataset

  • Title: AI in Social Media Statistics 2026 (Axis Intelligence Research dataset)
  • Creator: Axis Intelligence Research
  • Rows: 141 · Columns: row_id, metric, value, unit, as_of_date, geography, segment, source_org, source_document, source_url, retrieved_date, is_primary, axis_calculated, method_note
  • Coverage: marketer adoption, platform ad automation, AI labeling, audience perception, evidence quality, social software vendors; 2023–2026
  • Format: CSV, UTF-8, ISO 8601 dates, NA for missing values
  • License: CC BY 4.0. Reuse is free with attribution to Axis Intelligence Research and a link to this page.

Cite This Research

APA: Axis Intelligence Research. (2026, October 20). AI in social media statistics 2026: 94% of social teams use AI, 3 billion labeled TikToks and a 91-point disclosure gap. Axis Intelligence. https://axis-intelligence.com/ai-in-social-media-statistics/

MLA: Axis Intelligence Research. “AI in Social Media Statistics 2026: 94% of Social Teams Use AI, 3 Billion Labeled TikToks and a 91-Point Disclosure Gap.” Axis Intelligence, 20 Oct. 2026, axis-intelligence.com/ai-in-social-media-statistics/.

Chicago: Axis Intelligence Research. “AI in Social Media Statistics 2026: 94% of Social Teams Use AI, 3 Billion Labeled TikToks and a 91-Point Disclosure Gap.” Axis Intelligence, October 20, 2026. https://axis-intelligence.com/ai-in-social-media-statistics/.

FAQ

What share of Meta’s ad revenue now runs through Advantage+ AI?

At least 31.6%, by Axis Intelligence Research’s calculation. Meta reported an Advantage+ annual run rate above $75 billion on its Q2 2026 call; dividing by annualized Q2 2026 ad revenue ($59.363 billion × 4) gives 31.6%, up from 30.0% in Q3 2025.

How many videos has TikTok labeled as AI-generated?

Over 3 billion as of July 10, 2026, up from over 1.3 billion on November 19, 2025. That works out to roughly 7.3 million new labels a day, applied through C2PA Content Credentials, creator self-labels and invisible watermarking.

How much of a new TikTok account’s For You feed is AI content?

In Kapwing’s May 2026 test of one new account, 294 of 500 videos (59%) were classified as AI slop. Treat this as a cold-start snapshot: established accounts with viewing history see different recommendations, and Kapwing sells AI video tools.

Does AI-generated social content perform better than human-made posts?

There is no independent evidence either way yet. In Sociality.io’s small survey, 44.7% of professionals said AI-assisted posts perform better, 5.3% worse and 31.6% had not compared. These are self-reports from about 38 respondents, not measured engagement data.

What do users do when they spot AI slop from a brand?

Many disengage. Sprout Social found 50% of Gen Z and 29% of Boomers have unfollowed, muted or blocked an account over AI slop, and unlabeled AI content tops the list of things users want brands to stop posting (28%).

Which AI capabilities are social media teams underusing?

Listening and governance. Only 13.54% of social marketers use AI for social listening and trend monitoring, and 9.75% for brand voice auditing, against 42.9% for ideation, per HubSpot 2026 data.

Should brands label AI-generated social media posts?

The audience data points that way. Pew found 76% of US adults consider it important to distinguish AI from human content, and Sprout Social users rank unlabeled AI content as the top thing brands should stop. TikTok, Meta and YouTube also apply their own labels when they detect AI content.

Are social media marketers still using AI image generators?

Fewer than a year ago. HubSpot’s 2026 survey of 1,007 marketers shows visual AI tool use falling from 53% to 31% while chatbot use rose from 52% to 71%. Image generation is also where AI most often falls short (40.2%).

What is the Social AI Penetration Index (SAPI)?

The Social AI Penetration Index (SAPI) is a 0–100 composite by Axis Intelligence Research measuring how deeply AI has entered social media across creation, ad buying, disclosure and audience exposure, each weighted 25%. Its baseline reading is 46.1 as of October 20, 2026.

Is the widely shared “71% of social media images are AI-generated” statistic real?

We could not trace it to any primary source. The largest measured study we found, covering nearly 15 million Twitter profile photos in 2023, identified 0.052% as AI-generated faces. Use category-level measurements, such as Kapwing’s 2026 TikTok audit, instead.

Recent Posts

Data Center Opposition 2026-2027: 843 Groups, 120 Stalled Projects and 46 of 47 Local Votes

AI Data Center Opposition 2026-2027 By Axis Intelligence Research Co-author: Sarah Mitchell | Last updated: November 22,

GTA 6 Statistics 2026-2027: Sales, Players, Reviews and Records

GTA 6 Statistics 2026-2027 By Axis Intelligence Research Co-author: Noah Mc | Last updated: November 22, 2026 | License:

How Many Companies Use AI? 2026 Data, 39 Named Cases and the 2027 Outlook

How Many Companies Use AI? By Axis Intelligence Research Co-author: Sarah Mitchell | Last updated: November 22, 2026 | L

Axis Intelligence Research

Stay ahead on tech & data

Get notified when we publish or update datasets, trackers, research, and reports across technology, business, AI, cybersecurity, finance, infrastructure, energy, and more.

Research updates only. No spam. Unsubscribe anytime.