This preprint argues that the DESI DR2 phantom-divide crossing inferred from late-time reconstructions of w (z) can be an apparent effect produced by an irreversible Dark Energy → Dark Matter (DE→DM) energy transfer, rather than evidence for a physical phantom component. In standard analyses, dark energy and dark matter are assumed to be separately conserved (Q = 0). If an interaction (Q > 0) is present, the same background evolution can be mis-mapped into an effective equation of state with wₑff < −1, even when the underlying dark-energy state remains wₜrue = −1. Within the Cosmology of Time (CT) framework, we propose a physical origin for this interaction: a constrained scaling (“Bridge Equation”) derived from Landauer’s bound applied to the Bekenstein–Hawking entropy production rate of the apparent (Hubble) horizon. The interaction is (i) irreversible, fixing the direction of energy flow DE→DM, and (ii) kinematically suppressed as the Universe approaches the de Sitter limit. When modulated by a decoherence-efficiency kernel expected to peak near the deceleration–acceleration transition (zₐcc ≈ 0. 63), the drain becomes localized in the same redshift window where DESI typically places the apparent crossing (z ≈ 0. 4–0. 7). Falsifiable predictions include: a non-monotonic wₑff (z) morphology (a localized “phantom-like dip”), distinct from monotonic CPL trends; an association between the apparent crossing redshift and zₐcc; a localized modification to structure-growth observables (e. g. , fσ8 (z) ) near the transition epoch, to be tested via CLASS/CAMB implementation. A full likelihood-level confrontation with DESI DR2/DR3 and Euclid, including perturbation-level treatment, is identified as the next step. DESI DR2; phantom crossing; dark energy–dark matter interaction; interacting dark energy; w (z) ; Landauer principle; horizon entropy; Bekenstein–Hawking entropy; decoherence; dark degeneracy; growth of structure; fσ8; physical cosmology; arrow of time.
Everton Behenck (2026) studied this question.