The core assertion — that only one Gold Card Visa has been approved against a backdrop of $1.3 billion in claimed sales — is fully supported by multiple independent sources at Tier 4 and the official government website at Tier 1. Commerce Secretary Lutnick confirmed exactly one approval in sworn congressional testimony on April 23, 2026, and multiple outlets independently confirm he had previously claimed $1.3 billion 'worth' was sold within days of the December 2025 launch. The $1 million cost figure is accurate as the widely-reported headline price, though the official program structure specifies the $1 million is a 'gift' paid after vetting, with applicants initially paying only a $15,000 nonrefundable processing fee — meaning the program 'costs' $1,015,000 for approved applicants, not $1 million at the point of application. The claim's secondary characterization of the $1.3 billion as '$1.3 billion in applications' introduces a material imprecision: all authoritative sources (AP/NBC, Washington Times, IMI Daily, official program website) confirm Lutnick described the figure as $1.3 billion 'worth' sold or in 'commitments,' not a count of formal applications; the subsequent court filing revealed only 338 formal requests had been submitted in total, making 'applications' a substantively different and overstated characterization of what Lutnick actually claimed. Per the MOSTLY TRUE standard, the core directional assertion (one approval vs. massive claimed demand) is substantially correct and the imprecision — '$1.3 billion in applications' vs. '$1.3 billion worth sold/committed' — does not reverse the claim's directional meaning but is identifiable and non-trivial. A reasonable viewer would still agree with the overall point, but the 'applications' framing inflates the scale of the discrepancy by conflating a marketing/sales claim with a formal application count. The TRUE boundary is not crossed because the $1.3 billion characterization as 'applications' is a specific factual imprecision not fully supported by sources (per Section defining MOSTLY TRUE: 'core assertion substantially correct but contains one or more identifiable inaccuracies that do not reverse its directional meaning').

Methodology note: The TRUE/MOSTLY TRUE boundary question here turns on how precisely a broadcast summary must characterize an underlying primary-source claim. The claim is a paraphrase of Lutnick's own words, which themselves were ambiguous ('$1.3 billion worth' never specified whether this was formal applications, expressions of interest, or processing-fee payments). Because the claim's STAT designation invites precision evaluation, and because the subsequent court filing (338 requests) makes clear that '$1.3 billion in applications' is not an accurate description of the program's actual intake at any point, MOSTLY TRUE is the appropriate boundary placement. If verbatim_quote had been captured, this assessment might have distinguished more cleanly between Tapper paraphrasing Lutnick's ambiguous claim vs. independently asserting an application count.