Cyabra Finds Coordinated Inauthentic Activity in Five of Six Competitive Midterm Races, Affecting Candidates from Both Parties
Cyabra found about 1 in 4 candidate-related posts on TikTok came from inauthentic accounts, versus under 1 in 20 on Facebook.
Sentiment and the balance of points
Rhea-AI Sentiment reads the wording of the document, how positive or negative its language is on a 1 to 5 scale. The balance of points shown with the takes weighs what the document actually discloses, so the two can disagree, for example when a trial that missed its main goal is described in upbeat language.
Rhea-AI Summary
Cyabra (CYAB) released its 2026 U.S. Midterm Report examining online activity surrounding six competitive Senate and governor’s races. The analysis covered about 274,000 posts and comments across X, TikTok, Facebook and Instagram from August 31 to October 19. Cyabra identified coordinated inauthentic accounts with high confidence in five races and found about 1 in 9 posts about candidates came from inauthentic accounts, at roughly the same rate for Republican and Democratic candidates.
Cyabra and Carahsoft will host a December 9, 2026 luncheon where Cyabra plans to introduce a global protocol for collecting and analyzing online election narratives, drawing on its analysis of elections in five countries.
Key Figures
- Posts and comments analyzed
- About 274,000
- Across X, TikTok, Facebook and Instagram; no sampling
- Overall inauthentic-post share
- About 1 in 9 posts
- Posts about candidates; roughly the same rate for Republican and Democratic candidates
- Races with coordinated inauthentic accounts
- 5 of 6
- Identified with high confidence
- Races with attacks on both candidates
- 4 races
- Coordinated attacks hit both Republican and Democratic candidates
- Inauthentic comments on candidates' X accounts
- Between 1 in 3 and 1 in 5
- In most races; up to three times the overall rate
Historical Context
-
Council formation preceded this report's addition of election integrity to U.S. public-sector national-security work.
24h Move is the share-price change in the day after each event; other market factors may also have contributed.
Key Terms
open-source intelligence technical
AI-generated analysis. How Rhea-AI works. Not financial advice.
Cyabra’s 2026 U.S. Midterm Report finds about 1 in 9 posts about the candidates came from inauthentic accounts, at roughly the same rate for Republican and Democratic candidates
New York, NY, Oct. 28, 2026 (GLOBE NEWSWIRE) -- Cyabra, Inc. (Nasdaq: CYAB) (“Cyabra” or the “Company”), an artificial intelligence (“AI”)-powered narrative intelligence company that helps governments and enterprises detect coordinated manipulation and protect digital trust, today released Cyabra’s 2026 U.S. Midterm Report, an evidence-based narrative open-source intelligence analysis of inauthentic activity and coordination in online conversation. The report adds election integrity to Cyabra’s national security work in the U.S. public sector, following the Company’s formation earlier this month of its National Security Advisory Council. Cyabra will further discuss the report at a December 9, 2026 luncheon with Carahsoft for public-sector leaders and election stakeholders.
The report examined six of the most competitive races of the 2026 cycle, the U.S. Senate contests in Kansas, Maine and Michigan and the governor’s races in Georgia, Nevada and Ohio. Cyabra analyzed about 274,000 posts and comments across X, TikTok, Facebook and Instagram from August 31 to October 19, with no sampling. For each race, Cyabra identified the leading narratives about both candidates, measured the rate of inauthentic profiles and content behind them, assessed whether that activity was coordinated, and traced how specific narratives were amplified. The report found:
- About 1 in 9 posts about the candidates came from inauthentic accounts, at roughly the same rate for Republican and Democratic candidates. By platform, the share was about 1 in 4 on TikTok, 1 in 7 on X, 1 in 11 on Instagram and under 1 in 20 on Facebook.
- Between 1 in 3 and 1 in 5 comments on a candidate’s own X account came from inauthentic accounts in most races, up to three times the overall rate.
- In five of the six races, Cyabra identified coordinated inauthentic accounts with high confidence; in four, coordinated attacks hit both the Republican and Democratic candidates.
- Inauthentic activity peaked around the candidate debates in mid-October, and such activity was still active in Ohio and Michigan when Cyabra’s analysis window closed on October 19.
During an election, a coordinated attack can look like a groundswell, and inauthentic accounts can pass for real voters. For governments, protecting elections from foreign and domestic disinformation and safeguarding national sovereignty both depend on knowing what is authentic, whether activity is coordinated and, where the evidence allows, who is behind it. Cyabra’s methodology gives election agencies early warning of influence operations and coordinated disinformation, with confidence-rated findings they can act on before false narratives take hold. It also measures narrative reach, sentiment and how much engagement comes from authentic versus inauthentic accounts.
“Influence operations are now a weapon system and public trust is being shaped by forces most citizens can’t see,” said Dan Brahmy, Co-Founder and Chief Executive Officer of Cyabra. “We tracked the run-up to Election Day, so the picture is in hand while the cycle is still live, when knowing which activity is inauthentic and which activity is coordinated is most useful to the institutions responding to it. Our standard is the same for every candidate and every party. Evidence first, and conclusions only as far as the evidence goes.”
The next step comes on December 9, 2026, when Cyabra and Carahsoft will host a luncheon for public-sector leaders and election stakeholders, where Cyabra will introduce its global protocol for narrative open-source intelligence collection and analysis in elections. The protocol draws on Cyabra’s analysis of national elections in five countries and sets out how online political conversation is collected, how inauthentic and coordinated activity is identified and how findings are confidence-rated and reported. Additional details, including registration information, will be announced.
“Agencies need findings they can defend under scrutiny and brief up the chain,” said Keith E. Noble, Ph.D., VP, Public Sector North America, at Cyabra. “This report shows what that looks like in a live election. On December 9, 2026, we’ll introduce a global protocol, drawn from elections in five countries, that agencies can apply to their own elections.”
Cyabra is nonpartisan and every finding in the report is drawn from publicly available data. Both candidates in every race were assessed using the same method, platforms and time window. Races with little inauthentic activity are reported as plainly as races with a great deal. Where the report identifies coordination, it describes the behavior, states a confidence level and notes where attribution cannot be established. The report takes no position on any candidate, party, policy or electoral outcome.
The report builds on Cyabra’s recognition as a Market Shaper in the inaugural June 2026 Gartner® Emerging Market Quadrant for Narrative Intelligence - Startup Vendors, in which Gartner, Inc. formally defined and assessed the narrative intelligence market for the first time.¹
The full report and state-by-state analysis can be accessed here (https://cyabra.com/reports/coordinated-inauthentic-activity-in-the-2026-us-midterm-races/).
About Cyabra
Cyabra is an AI-powered narrative intelligence company that helps national security and defense organizations, government agencies, brands, communications agencies, and global enterprises restore trust and authenticity online by analyzing manipulated content, coordinated behaviors, and inauthentic actors. The platform helps teams understand who is operating, how activity is amplified, and where coordinated activity is shaping perception, translating evidence into clear mitigation steps. By reducing ambiguity and misdirected response, Cyabra enables proportionate, evidence-led action when clarity matters most.
For more information, visit www.cyabra.com.
Contact:
Investors: ir@cyabra.com
Media: pr@cyabra.com
Forward-Looking Statements
This press release contains forward-looking statements within the meaning of the United States Private Securities Litigation Reform Act of 1995. Forward-looking statements include all statements that are not historical statements of fact and statements regarding Cyabra’s intent, belief, or expectations, including, but not limited to, statements regarding Cyabra’s future results of operations and financial position, planned products and services, business strategy and plans, market size and growth opportunities, competitive position and market trends. Some of these forward-looking statements can be identified by the use of forward-looking words, including “may,” “should,” “expect,” “intend,” “will,” “estimate,” “anticipate,” “believe,” “predict,” “plan,” “targets,” “projects,” “could,” “would,” “continue,” “forecast” or the negatives of these terms or variations of them or similar expressions. For example, the Company is using forward-looking statements in this press release when discussing the planned December 9, 2026 event and the anticipated introduction of the global elections protocol, the anticipated uses and benefits of the report and the protocol for public-sector organizations, potential benefits of the National Security Advisory Council, the Company’s plans to continue analyzing election-related online activity and publishing related findings and the Company’s expansion in the U.S. public sector. These statements relate to future events and involve known and unknown risks, uncertainties, and other factors which may cause actual results, performance or achievements to be materially different from any future results, performance or achievements expressed or implied by the forward-looking statements. Such factors include, among others, that the planned December 9, 2026 event may be rescheduled or may not occur, that the anticipated benefits and uses of the report and protocol may not be realized, that the Company’s findings may be affected by limitations in available data, methodology or attribution, and those set forth in Cyabra’s filings with the Securities and Exchange Commission. Prospective investors are cautioned not to place undue reliance on such forward-looking statements, which speak only as of the date of this press release. Cyabra undertakes no obligation to publicly update any forward-looking statement, whether as a result of new information, future events or otherwise.
_______________________
1 Source: Gartner, Emerging Market Quadrant for Narrative Intelligence – Startup Vendors, Alfredo Ramirez IV, Apeksha Kaushik, Akif Khan, Amber Boyes and David Senf, June 26, 2026. GARTNER is a registered trademark and service mark of Gartner, Inc. and/or its affiliates in the United States and internationally and is used herein with permission. All rights reserved. Gartner does not endorse any vendor, product or service depicted in its research publications and does not advise technology users to select only those vendors with the highest ratings or other designation. Gartner publications consist of the opinions of Gartner’s business and technology insights organization and should not be construed as statements of fact. Gartner disclaims all warranties, express or implied, with respect to this research, including any warranties of merchantability or fitness for a particular purpose.
FAQ
AI-generated questions and answers. How Rhea-AI works. Not financial advice.
What did Cyabra’s 2026 U.S. Midterm Report find?
Cyabra found that about 1 in 9 posts about candidates came from inauthentic accounts, at roughly the same rate for Republican and Democratic candidates. It identified coordinated inauthentic accounts with high confidence in five of the six races examined.
How did Cyabra analyze online activity for its 2026 midterm report?
Cyabra assessed both candidates in every race using the same method, platforms and time window, with no sampling. All findings used publicly available data. Where the analysis identified coordination, it described the behavior, stated a confidence level and noted where attribution could not be established.