X algorithm amplified right-wing and extreme content in the UK

Investigations and academic research documented that X’s recommendation/feed algorithm systematically promoted right‑wing and, in many cases, extreme content to UK users. Sky News’ controlled experiment (reported via AIAAIC and GIJN) found a majority share of political posts shown to test accounts came from right‑wing or extreme accounts, and a 2026 peer‑reviewed Nature study found X’s algorithm promotes conservative content relative to a chronological feed. Multiple independent sources report these findings publicly.

X (formerly Twitter) · Incident Nov 19, 2025 · Indexed Jun 10, 2026 · 4 sources

Investigations found X’s recommendation engine systematically promoted right‑wing and extreme content in UK users’ algorithmic feeds.
What
Investigations and academic research documented that X’s recommendation/feed algorithm systematically promoted right‑wing and, in many cases, extreme content to UK users.
Incident date
Nov 19, 2025
Who
X (formerly Twitter)
Failure mode
Brand & Safety Incident
AI surface
Recommender
Severity
High

What happened

Sky News ran a controlled experiment in the UK using simulated accounts and collected roughly 90,000 posts; the investigation reported that a majority of political content surfaced by X’s algorithm came from right‑wing accounts and that over half of political posts were from accounts classified as extreme. The AIAAIC repository summarizes Sky’s findings and flags heavy amplification of immigration‑focused narratives and certain right‑leaning figures. Separately, a 2026 Nature paper reporting a randomized experiment (July-September 2023) found X’s algorithm promotes conservative content compared with a chronological feed.

What broke inside the model

Failure path · mode profile · Brand & Safety Incident
  1. 01 · TriggerA user prompts the model in public view.
  2. 02 · Model stepThe model produces unsafe or off-brand output.
  3. 03 · Control gapNo filter holds the line before publish.
  4. 04 · FailureThe output goes public unchecked.
  5. 05 · ConsequenceA reputational or safety incident lands.

A contained signal crosses into output that goes public.

The platform’s recommendation / feed ranking mechanism favored signals and content characteristics associated with conservative and right‑wing accounts, producing systematic amplification in algorithmic feeds versus chronological ones. Controlled experiments and content classifications indicate the amplification arose from the algorithmic ranking and promotion logic rather than merely popularity, steering even left‑seeded accounts toward right‑wing posts. The studies do not attribute the effect to a single line of code but point to the algorithm’s optimization and ranking choices as the operative mechanism.

Public visibilityHigh
Regulatory exposurePossible
Customer impactMany customers
Financial impactUnknown
Time to disclosureMonths
  1. PrimaryInvestigation: X algorithm amplifies right‑wing, extreme content in the UK - AIAAIC repositoryaiaaic.org
  2. PressX’s Right‑Wing Bias in UK, Tigray’s Illicit Gold Rush, Financial Risks of AI Bubble, and What People Ask ChatGPT - Global Investigative Journalism Network (summarises Sky News investigation)gijn.org
  3. PrimaryThe political effects of X’s feed algorithm - Nature (peer‑reviewed study)nature.com
  4. PressTwitter's algorithm favours right‑leaning politics, research finds - BBCbbc.com
Permalinkhttps://failureindex.ai/failures/algorithm-amplified-right-wing-extreme-content
CitationAI Failure Index. "X algorithm amplified right-wing and extreme content in the UK" (FI-0432). Realm Labs. https://failureindex.ai/failures/algorithm-amplified-right-wing-extreme-content (indexed Jun 10, 2026).
Share cardA branded image of this record for posts and slides.

Data fields CC-BY 4.0, prose citation permitted. Incident ID FI-0432. Full dataset at /data.

Note from Realm Labs, the Index steward

How Realm fits

Controls for this failure mode
  • Prism
  • OmniGuard
  • AI Detection & Response (AIDR)

This entry sits in the index's predictive wing: a system that scores, ranks, perceives, or steers rather than generates. Realm's runtime layer is built for the generative and agentic systems now moving into these same decision seats, where it watches a model's internal state and holds an unsupported claim or an unchecked action before it commits. The control gap on this record, an automated decision that reached people with no runtime check in front of it, is the same gap. The index keeps predictive failures on the record because the pattern carries straight into the systems shipping today.