TeV gamma rays from distant blazars arrive in greater numbers than the intervening light should permit. Very-high-energy photons traversing cosmological distances pair-produce on the extragalactic background light (EBL), the accumulated starlight and dust emission of all cosmic history, so spectra of distant TeV sources should show strong, redshift-dependent absorption. Multiple analyses have reported the opposite tendency: distant blazar spectra appear harder than absorption models allow, with the anomaly quantified as a lower-than-expected opacity at the highest energies and redshifts (Horns and Meyer 2012, claiming 4 sigma; subsequent studies disputing significance after EBL model and selection effects). The proposed exotic fix is axion-like particles: photons converting to ALPs in magnetic fields, traversing the EBL unscathed, and converting back, effectively making the universe translucent.
The dispute is genuinely unresolved: modern EBL models sit near the lower bounds allowed by galaxy counts, shrinking the room the anomaly needs, while spectral hardening claims survive in some source samples and evaporate in others depending on intrinsic-spectrum assumptions; CTA-precursor measurements keep both readings alive. The tension also couples awkwardly to its neighbor anomaly: New Horizons measurements of the cosmic optical background suggested more diffuse light than galaxy counts integrate, which would deepen absorption, the opposite direction the transparency anomaly wants, so the two backgrounds-and-absorption datasets pull against each other within the standard accounting.
The standing is a disputed anomaly awaiting its instrument: CTA's order-of-magnitude sensitivity gain will measure blazar spectra and EBL opacity precisely enough to confirm the transparency excess or retire it, and to discriminate ALP conversion signatures from EBL mismodeling.