Andrej Karpathy's 19 May 2026 announcement that he has joined Anthropic, paired with Anthropic's confirmed framing that he will lead a team focused on using Claude to accelerate pre-training research (under team lead Nick Joseph), is a foundational-layer vendor-trajectory signal that composes with the 5 May 2026 Wall Street agents launch (AM-159) to describe Anthropic operating on both ends of the platform stack simultaneously — vertical-depth-first on the application layer and name-recognition-first on the pre-training layer. The mandate (Claude accelerating Claude) is more procurement-relevant than the hire itself, because it is a public commitment to recursive self-improvement of the model line at the foundational layer rather than at the application layer. By 17 August 2026, observable evidence in the AI-research community will or will not appear across four markers: (1) a published paper from Anthropic's pre-training team describing a Claude-in-the-loop component with measurable productivity or capability impact; (2) a Claude release crediting Claude-assisted research methodology in the development cycle; (3) public commentary from Karpathy or Anthropic leadership on team progress beyond the launch-day framing; (4) Anthropic-attributed performance gains on community-authoritative benchmarks. Procurement-template implication: AI-vendor questionnaires should add a model-improvement-methodology disclosure field, and multi-year MSAs should add a research-roadmap-attestation clause requiring thirty-day advance notice on material methodology changes.

Claim is scoped to enterprise CIOs sizing multi-year AI-platform commitments against the May 2026 vendor-trajectory evidence from the four-lab cohort (Anthropic, OpenAI, Google, Microsoft). 90-day review cadence calibrated to the time required for the first observable markers of operational follow-through to appear. The launch-day framing is necessarily aspirational; the review window tests whether foundational signs of operational follow-through emerge, not whether the team has shipped finished output. Trigger conditions for status changes: (1) by 17 Aug 2026, presence of one or more of the four observable markers (research paper; release with credited methodology; commentary; attributed benchmark gains) would harden the recursive-self-improvement reading and keep Holding; (2) absence of all four would move toward Partial because the launch-day framing was aspirational without operational follow-through; (3) Karpathy publicly departing Anthropic before the 17 Aug review would move toward Not holding on the strong reading because the name-recognition signal was retracted; (4) a comparable hire at OpenAI, Google DeepMind, or Meta announced with a comparable model-self-improvement mandate inside the review window would broaden the pattern from Anthropic-specific to industry-wide, refining the audience-scope but not the load-bearing claim; (5) a Microsoft-publicly-disclosed expansion of model-assisted-research investment at OpenAI inside the review window would refine the Anthropic-vs-OpenAI vendor-trajectory delta and could move toward Partial on the strong vendor-distinction reading.