The arXiv preprint "AutoResearch AI: Towards AI-Powered Research Automation for Scientific Discovery" (Tie et al., 2026) surveys systems that integrate literature grounding, hypothesis generation, experimentation, validation, reporting, and revision into longer-horizon workflows. Five workflow conditions are defined: literature and research grounding; hypothesis formation and planning; experimentation and tool use; feedback, validation, and review; and reporting and knowledge communication, with systems redistributing control across evidence, execution, and accountability. Vibe Research is positioned as human-steered prompt assistance while AI-led coordination remains short of robust autonomy. The survey synthesizes AI scientist systems, mixed-initiative frameworks, benchmarks, and domain deployments while identifying gaps in evidence preservation, reproducibility, provenance tracking, and cross-domain robustness. Five evaluation dimensions are proposed—novelty, validity, impact, reliability, and provenance—with autonomy shown as domain-conditioned and more credible in structured, executable, rapidly verifiable settings. Related work in arXiv:2308.09662 documents parallel patterns in autonomous hypothesis iteration and tool invocation.