Will SpaceX demonstrate broadband direct-to-device service (>5 Mbps downlink to unmodified phone) by December 31, 2026?

Q4 2025 earnings produced no new SpaceX D2D data — the most informative update possible was silence. The physics analysis is unchanged: consumer phones have fixed, low-power antennas requiring very large satellite-side aperture to close the broadband link budget. SpaceX's 600+ D2D fleet is confirmed narrowband-optimized hardware and cannot be software-upgraded to broadband. The remaining ~10 months until December 31, 2026 is marginally shorter than at the previous prediction date, which fractionally reduces probability. No positive SpaceX evidence has emerged, and the CRITICAL data gap on SpaceX's roadmap — the primary source of residual probability — is unchanged. The physics-constrained base case is firmly NO. AST's strengthening ecosystem (new MNO partners, ASIC chip, manufacturing scale) does not directly affect SpaceX's probability but confirms the technological differentiation thesis.

SpaceX's extraordinary iteration speed is the primary counterweight to the physics constraints. The CRITICAL data gap on SpaceX's satellite design roadmap remains unresolved — stealth development of broadband-capable D2D hardware cannot be excluded. The multi-satellite beam-coordination path to >5 Mbps without per-satellite large aperture is theoretically viable and SpaceX has demonstrated sophisticated coordination capability across large constellation fleets. However, the Q4 earnings cycle produced no hint of SpaceX broadband D2D progress from any industry participant, including AST management who has strong incentive to surface competitive intelligence. The T-Mobile partnership scope remains messaging-only. Given these offsets, I maintain a slightly elevated probability to reflect engineering uncertainty and the CRITICAL data gap, but the marginal time compression nudges the estimate down slightly from my prior of 0.18.

Multi-satellite beam concentration remains SpaceX's most plausible technical path to >5 Mbps without redesigning individual satellites. This approach — coordinating multiple narrowband-optimized satellites to concentrate energy on a single handset — would not require new satellite hardware generation and could theoretically be demonstrated within the ~10 month window. However, the uplink bottleneck from the consumer phone's fixed antenna gain is an independent physics constraint that beam concentration from the space segment cannot solve. Resolution requires a two-way data session exceeding 5 Mbps downlink to an unmodified phone, and without a corresponding uplink capability, this remains an asymmetric challenge. The Q4 earnings update adds no positive evidence; the symmetry of the evidence base supports the prior estimate with slight downward pressure from time compression.

Weighing the structured evidence: three independent lenses confirm SpaceX broadband D2D is unproven; the physics analysis identifies large satellite aperture as the binding constraint; the current SpaceX fleet is narrowband-only hardware; and the Q4 2025 earnings cycle produced zero new SpaceX broadband D2D evidence from any source. The CRITICAL data gap on SpaceX's design roadmap prevents assigning very low probability, but absence of evidence over a 22-day period including a major earnings cycle is mildly negative. The AST CTO's 94.4% divestiture remains an ambiguous signal — the person most likely to understand the competitive dynamics has nearly fully exited, but this could reflect personal financial planning rather than competitive pessimism. The marginal time compression (10 months → ~10 months) is a small but real factor.

The resolution criterion requires a PUBLIC demonstration — internal capability is insufficient. As of March 2, 2026 there is no public evidence of SpaceX broadband D2D development, no T-Mobile announcement of broadband D2D roadmap, and no FCC filings indicating broadband D2D testing. The 22 days since the previous prediction batch have passed with no new information, which constitutes weak but real negative evidence. The commercial incentive structure (SpaceX and T-Mobile would benefit enormously from a broadband D2D demonstration) combined with continued silence suggests the capability does not yet exist. The CRITICAL data gap prevents going below 10%, but the overall evidence weight points solidly toward NO.

Monitoring trigger context: this market tests the single most critical competitive escalation trigger for AST — a SpaceX broadband D2D demonstration would collapse AST's primary differentiation and likely reprice the stock. Trigger probability should reflect both the physics constraints and the possibility that SpaceX is deliberately sequencing capabilities (text → voice → data) along a roadmap that could reach broadband before year-end. The FCC SCS license granted December 2025 enables commercial broadband service; SpaceX now has regulatory runway if hardware capability follows. However, regulatory runway does not create hardware capability. The key question is whether SpaceX's next-generation D2D satellite design (which may or may not exist) includes larger antenna aperture. Without intelligence on that design, I assign probability consistent with the physics-constrained base case plus a technology uncertainty premium.

SpaceX D2D fleet is narrowband-only. Broadband to unmodified phones requires different satellite hardware. Q4 2025 earnings had no SpaceX broadband news. ~10 months remaining is still too short for hardware redesign, launch, and public demo. Slightly lower than prior 0.12 estimate due to passing time and absence of positive evidence.

Satellite generation leaps take 2-4 years historically. Narrowband to broadband D2D is a major capability jump. The window has now compressed to ~10 months. No positive evidence has emerged from Q4 2025 earnings or any other channel. The base rate for this kind of capability leap in this timeframe is low, and passing time without evidence is a mild negative signal.

SpaceX currently cannot demonstrate broadband D2D. The CRITICAL data gap on stealth development means very low probability is not appropriate. But all evidence — physics constraints, Q4 earnings silence, no T-Mobile broadband announcements — points to NO. The base case remains unchanged from the February 8 prediction. Time compression is a marginal factor.