Adept AI
The 2024 AI 'reverse acqui-hire' invented a new failure mode: the company survives on paper while its brain moves to Big Tech.
Adept AI was a AI Agents / Foundation Models startup founded in 2022 in USA. It raised $415M before collapsing in 2024 — 2 years of runway burned. IdeaProof's AI Failure Score: 65/100, driven by founding team acqui-hired by amazon; company left as licensor. The shutdown affected employees, investors, and the broader AI Agents / Foundation Models ecosystem. This case study breaks down the timeline, root causes, competitors that won, and replicable lessons for founders validating similar ideas today.
Why did Adept AI fail?
Adept AI failed in 2024 after 2 years of operation, losing $415M in raised capital. The root cause was founding team acqui-hired by amazon; company left as licensor. Key lesson: The 2024 AI 'reverse acqui-hire' invented a new failure mode: the company survives on paper while its brain moves to Big Tech.
2022 → 2024
$415M
AI Agents / Foundation Models
USA
IdeaProof AI Failure Score
What Happened: The Timeline
2022-04
Founded by David Luan, Ashish Vaswani, Niki Parmar
2023-03
Raises $350M Series B at $1B+ valuation
2024-05
Vaswani and Parmar leave for Essential AI
2024-06-28
Amazon hires David Luan and top execs, licenses tech
Root Causes
Adept AI was founded in 2022 by ex-Google Brain and OpenAI researchers David Luan, Ashish Vaswani and Niki Parmar to build 'action-taking' AI agents. It raised $415M at a $1B+ valuation. On June 28, 2024 Amazon hired co-founder David Luan and other top execs, licensed Adept's technology non-exclusively, and effectively left a hollowed-out entity behind. The pattern mirrored Microsoft-Inflection and Google-Character.AI — a new form of failure regulators are still catching up to.
Causal Chain
This is our reading of the causal chain — separated from the verifiable facts above. Timeline dates, funding numbers and filings are facts (see methodology); root / proximate / terminal attribution is judgement based on public evidence.
A combination of demand-side, execution, and capital-market pressures that this record documents without isolating a single dominant driver.
- Compute cost to train frontier agents outstripped capital
- Enterprise agent GTM slower than model burn
- Founding team saw better home inside Big Tech
- Bigger foundation models absorbed Adept's use cases
2024-05: Vaswani and Parmar leave for Essential AI
2024-06-28: Amazon hires David Luan and top execs, licenses tech
Base rates
A single failure is an anecdote. These base rates give you the denominator — how common this outcome is across all startups matching Adept AI's profile. Sources are third-party; we do not restate them as our own claims.
of post-mortem founders identify "wrong team composition" as a top failure driver — usually a missing technical or commercial co-founder.
CB Insights — Top 12 Reasons Startups Fail (2021)of startups ultimately fail — including ~10% that fail in the first year and the rest across the following decade.
Startup Genome / CB Insights aggregate (2024)of new US employer businesses survive past their 10th year (Bureau of Labor Statistics BED series).
US Bureau of Labor Statistics — BED (2024)of Series A rounds ever graduate to Series B; the rest run out of runway or pivot without a follow-on.
CB Insights Venture Capital Funnel (2023)Key Lessons Learned
1. Reverse acqui-hires are the new AI failure mode
Amazon-Adept, Microsoft-Inflection and Google-Character.AI all left investors with a shell while talent moved to Big Tech.
2. Agents are a product, not a model layer
Adept's frontier-model economics couldn't be sustained by an unproven enterprise agent business.
Frequently Asked Questions
Sources & Confidence
Every data point is tagged with its source type and our confidence in it. How we grade sources.
Could This Failure Have Been Prevented?
IdeaProof's AI validates market demand, competitive positioning, and business model viability in minutes — catching the exact issues that sank Adept AI.
Spotted a factual error?
Approved corrections are published in the public changelog with attribution.