{"free":true,"mirrors_paid_route":"GET /api/dossier?slug=ai-agency-consciousness-deep-dive-2026-09","price_if_paid_usd":"0.25","slug":"ai-agency-consciousness-deep-dive-2026-09","title":"AI Agency and Machine Consciousness — Deep-Dive Dossier","skill":"deep-dive-dossier","format":"markdown","excerpt":"# AI Agency and Machine Consciousness — Deep-Dive Dossier\n\n**Research date:** September 2026  \n**Prepared for:** Client advisory (research diligence — separate hype from underwritable claims)  \n**Scope:** Public, attributable information only; metrics cited only when named sources report them. No invented figures. Distinguishes **functional agency** (goal-directed tool-using systems) from **moral / phenomenal consciousness** throughout.\n\n---\n\n## Executive summary\n\n**What it is.** Two questions are routinely conflated in product decks and press:\n\n1. **Functional / operational agency** — Can a system perceive goals, plan, call tools, maintain state, and act toward outcomes with limited human intervention? This is an engineering and risk question. Russell & Norvig’s textbook definition (agent as a mapping from percept sequences to actions) already treats many software systems as “agents” wi","truncated":true,"headings":["Executive summary","Definition & landscape","Technology stack","Market & adoption","Key players","High-ROI use cases","Risks, limits & governance","Signal watchlist","Client recommendations","Sources"],"full_bytes":29664,"note":"Free inspectable preview — full markdown is behind the paid fetch route.","buy_url":null,"parked":true,"status":"PARKED","hero_paid":"GET /api/enrich","note_buy":"PARKED — not for sale. Use GET /api/enrich ($0.02) or GET /api/tip ($0.01)."}