Fact check: What percentage of news from left-wing and right-wing sources is factually correct according to independent fact-checkers?

1. Why the simple percentage claim collapses under scrutiny — the measurement problem

Independent fact‑checking requires operational definitions of “news,” “factually correct,” and a representative sample of left and right outlets; the supplied materials demonstrate that such definitions are uneven and often proprietary. Ground News claims to rate outlet factuality using AI and editorial methods, but its factuality ratings are paywalled, preventing independent verification of sampling and scoring choices ^[1]. The Pew methodology note shows how survey design matters for representative measurement, underscoring that methodological differences produce divergent accuracy estimates and complicate any single percentage claim ^[4].

2. What Ground News contributes — transparency gaps and potential utility

Ground News positions itself as a bias‑cutting aggregator that also assigns factuality ratings, and a 2024 study tied to the platform suggests cross‑partisan exposure can reduce polarization, which indirectly affects perceptions of accuracy ^[1]. However, the critical caveat is that Ground News’ detailed rating algorithm and outlet scores are behind a paywall, which limits external audit and raises questions about sample selection, weighting, and label calibration. This combination of proprietary scoring and claims of neutrality creates an information asymmetry where the platform can influence discourse without subjecting its accuracy claims to public scrutiny ^[1].

3. Nonpartisan fact‑checking efforts — scope but not exhaustive percentages

Get Fact is presented as a non‑partisan, collaborative fact‑checking collective using human and machine methods to verify Canadian claims; it emphasizes sourcing transparency but does not provide a landscape‑level percentage comparing left and right outlet accuracy ^[2]. Fact‑checking organizations typically publish itemized verdicts on claims rather than aggregated accuracy percentages by political orientation, because aggregated metrics risk masking heterogeneity across topics, formats, and temporal contexts. The absence of an aggregate percentage in these summaries reflects a deliberate focus on case‑level verification rather than cross‑ideological quantification ^[2].

4. Case studies and claim‑level findings — accuracy varies by story, not simple axes

A fact‑check of a major political speech (Donald Trump’s UN address) found multiple false or misleading claims, illustrating how claim‑level checks reveal granular errors irrespective of outlet ideology ^[3]. Fact‑checkers typically catalog specific falsehoods, partial truths, and context omissions rather than rate entire outlets wholesale. This approach produces a public record showing that accuracy is story‑dependent; outlets on both the left and right can publish accurate reporting, errors, or misleading framings, which undermines the utility of a single percent figure for each side ^[3].

5. Structural and political factors that skew any simple comparison

Media consolidation, ownership changes, and platform dynamics shape what reaches audiences and how accuracy is enforced; the discussion of potential consolidation following high‑profile acquisitions signals that structural shifts can alter editorial incentives and fact‑checking pressures ^[5]. When ownership concentrates, editorial lines and resource allocation for fact‑checking may shift, complicating longitudinal comparisons between left and right sources. These systemic drivers mean that any percent figure would be highly sensitive to temporal, ownership, and platform context, limiting its generalizability ^[5].

6. Practical implications for evaluating news accuracy — what readers should expect

Because the provided sources lack an agreed numeric split, readers should rely on claim‑by‑claim verification, cross‑source triangulation, and scrutiny of fact‑checker methods rather than headline percentages. Aggregators and collaborative fact‑checkers provide valuable tools, but transparency about methodology and open access to ratings are essential for trustworthy aggregation; paywalls and opaque algorithms impede public assessment of claims about "what percentage" of left versus right reporting is accurate ^{[1] [2]}. The evidence suggests that accuracy is conditional and measurable only with rigorous, open methods.

7. Bottom line and where to look next for a defensible number

No defensible, independently verifiable percentage emerges from the supplied materials. To produce such a number would require: a clearly defined sampling frame of outlets, open access to fact‑checker coding and criteria, and peer review or replication of results. Absent that, the available evidence supports the conclusion that accuracy varies by claim and outlet, and that current public tools either do not publish aggregate percentages or do so behind paywalls, leaving the question unanswered by the provided sources ^{[1] [2] [3] [4] [5] [6]}.