The AI Graveyard
Where billion-dollar AI dreams went to die. Every failure is a lesson. Read them all.
Last updated: April 2026
Total Shut Down
52
Companies that failed or were acqui-hired
Total Raised & Lost
$14.8B
Capital destroyed or absorbed
Top Failure Cause
Open-source chatbot absorbed into LMSYS Org. Never had independent funding or sustainability plan.
Most common reason companies failed
By the Numbers
Common Failure Reasons
The Fallen
In Memoriam
Vicuna
Foundation Models† 2023What Happened
Open-source chatbot absorbed into LMSYS Org. Never had independent funding or sustainability plan.
Lesson Learned
Open-source projects need sustainable funding models. Community goodwill doesn't pay server bills.
Jasper AI
AI Writing† 2024$125M raisedWhat Happened
ChatGPT launch commoditized AI writing overnight. Jasper's core product became a free feature.
Lesson Learned
Don't build a feature, build a platform. When the foundation model is free, wrappers die.
Kustomer
Enterprise AI† 2023$174M raisedWhat Happened
Acquired by Meta for $1B, then deprioritized and effectively shut down as Meta refocused.
Lesson Learned
Acquisition isn't always success. Big tech acquirers can shelve your product overnight.
Arthur AI
AI Infrastructure† 2024$60M raisedWhat Happened
AI observability market was too early. Enterprises weren't deploying enough AI to need monitoring.
Lesson Learned
Monitoring AI needs mature AI adoption first. You can't sell seatbelts before people have cars.
Neeva
AI Search† 2023$77.5M raisedWhat Happened
AI-powered search engine couldn't compete with Google's distribution. Shut down and team joined Snowflake.
Lesson Learned
Search is a monopoly for a reason. Distribution beats technology every time.
Argo AI
AI Robotics† 2022$3.6B raisedWhat Happened
Ford and VW pulled funding simultaneously. $3.6B wasn't enough for autonomous driving.
Lesson Learned
Autonomous driving needs patient capital measured in decades, not quarters.
Mighty AI
AI Data† 2019$27M raisedWhat Happened
Data labeling platform acquired by Uber for autonomous driving data needs.
Lesson Learned
Data labeling became a commodity. Scale AI won, everyone else got acqui-hired.
Clarifai
Computer Vision† 2024$100M raisedWhat Happened
General computer vision platform struggled to compete as cloud providers built CV into their APIs.
Lesson Learned
General CV lost to specialized solutions. AWS, Google, Azure gave it away for free.
Anki
AI Robotics† 2019$200M raisedWhat Happened
Couldn't scale consumer robotics. Cozmo was beloved but margins were razor-thin.
Lesson Learned
Consumer hardware + AI is brutal. High R&D, low margins, fickle customers.
Jibo
AI Robotics† 2018$73M raisedWhat Happened
No product-market fit for social robots. Consumers wanted utility, not companionship from a robot.
Lesson Learned
AI needs utility, not personality. Cute doesn't convert to recurring revenue.
Rethink Robotics
AI Robotics† 2018$150M raisedWhat Happened
Collaborative robots Baxter and Sawyer were too slow for industrial use cases.
Lesson Learned
Factories need speed and precision, not intelligence. Industrial buyers won't sacrifice throughput.
Digital Reasoning
AI Finance† 2021$100M raisedWhat Happened
AI analytics for financial compliance acquired by Smarsh. Couldn't scale independently.
Lesson Learned
Enterprise AI needs distribution. Great tech without sales channels gets absorbed.
Babylon Health
AI Healthcare† 2023$1.2B raisedWhat Happened
Ran out of cash despite $1.2B in funding. SPAC listing couldn't save it.
Lesson Learned
Healthcare AI burns cash fast. Regulation, trust, and unit economics are brutal.
Olive AI
AI Healthcare† 2023$850M raisedWhat Happened
Overexpanded into too many healthcare verticals simultaneously. Mass layoffs preceded shutdown.
Lesson Learned
Horizontal plays in healthcare fail. Pick one workflow and dominate it.
IronNet Cybersecurity
AI Security† 2023$400M raisedWhat Happened
Went public via SPAC, then filed for bankruptcy. Revenue never matched the hype.
Lesson Learned
SPAC doesn't mean product-market fit. Public markets expose weak fundamentals fast.
Veritone
Enterprise AI† 2024$200M raisedWhat Happened
AI media analytics platform saw stock crash 90%+. Revenue growth stalled despite AI hype cycle.
Lesson Learned
AI hype doesn't sustain revenue. Wall Street eventually demands real margins.
x.ai
AI Agents† 2021$44M raisedWhat Happened
AI scheduling assistant acquired by Bizzabo. Single-feature AI couldn't sustain a company.
Lesson Learned
Single-feature AI tools get absorbed. Calendly won with simplicity, not AI.
Geometric Intelligence
Foundation Models† 2016What Happened
AI reasoning startup acquired by Uber to build their AI research division.
Lesson Learned
Fundamental AI research attracts acquirers. Uber bought a brain trust.
Element AI
Enterprise AI† 2020$257M raisedWhat Happened
Founded by Yoshua Bengio. Acquired by ServiceNow for a fraction of its peak valuation.
Lesson Learned
Star founders don't guarantee product-market fit. Research brilliance doesn't equal revenue.
Lilt
NLP† 2024$92M raisedWhat Happened
AI translation platform pivoted heavily and had layoffs. DeepL and Google Translate dominated.
Lesson Learned
Translation is a winner-take-all market. Free tools from Google killed paid alternatives.
Prisma Labs
AI Image† 2022$10M raisedWhat Happened
Prisma Labs created a viral AI photo filter app that peaked at millions of downloads but could not sustain user engagement or monetize effectively once the novelty of artistic style transfer filters wore off.
Lesson Learned
Consumer AI apps built on a single novelty feature face brutal retention challenges. Without continuous innovation, recurring value, and strong monetization mechanics, viral download numbers cannot sustain a company.
Apprente
NLP† 2021$7M raisedWhat Happened
Apprente built AI-powered voice ordering technology for drive-through restaurants. McDonald acquired the startup in 2019 but later sold the technology to IBM in 2021 after the system struggled with accuracy in noisy real-world environments.
Lesson Learned
Voice ordering AI could not reliably handle noisy drive-through environments, diverse accents, and complex menu modifications. McDonald acquisition and subsequent IBM sale confirmed the technology was not production-ready at scale.
Adept AI
AI Agents† 2024$415M raisedWhat Happened
Adept AI failed to ship a competitive product despite raising $415M in venture funding. Amazon acqui-hired the core research team in mid-2024, absorbing key talent into its AGI division rather than acquiring the company outright.
Lesson Learned
Agents were ahead of their time in 2022. Timing matters as much as technology.
Blue River Technology
AI Robotics† 2017$30M raisedWhat Happened
Acquired by John Deere for $305M. AI-powered precision agriculture proved valuable to incumbents.
Lesson Learned
Agtech AI has real value. Sometimes a $305M acquisition on $30M raised is a win.
Bonsai
AI Infrastructure† 2018$36M raisedWhat Happened
Acquired by Microsoft. Reinforcement learning for industrial control found a home in Azure.
Lesson Learned
Deep tech AI gets acqui-hired by cloud giants. Microsoft paid for the team and the tech.
Maluuba
NLP† 2017$11M raisedWhat Happened
NLP research lab acquired by Microsoft to bolster AI research capabilities.
Lesson Learned
Research labs get acqui-hired. Pure research needs a patron with deep pockets.
DeepScale
AI Robotics† 2019$18M raisedWhat Happened
Autonomous driving perception startup acquired by Tesla for its engineering talent.
Lesson Learned
In AI, talent is the real product. Tesla bought the team, not the tech.
Voicera
Enterprise AI† 2019What Happened
AI meeting notes tool acquired by Cisco. Meeting AI was a niche feature in 2019.
Lesson Learned
Meeting AI was too early in 2019. Otter.ai survived, but most didn't.
Nervana Systems
AI Infrastructure† 2016$24M raisedWhat Happened
AI chip startup acquired by Intel for $400M. Purpose-built AI silicon attracted hardware giants.
Lesson Learned
AI chip startups have exit paths. Intel paid 16x raised for custom silicon expertise.
Sentient Technologies
Enterprise AI† 2019$143M raisedWhat Happened
AI-driven e-commerce optimization and hedge fund. Pivoted multiple times, never found product-market fit.
Lesson Learned
Raising $143M doesn't buy product-market fit. Focus beats funding every time.
Butterfly Network
AI Healthcare† 2024$350M raisedWhat Happened
AI-powered handheld ultrasound went public via SPAC, stock collapsed 90%+.
Lesson Learned
Medical device + AI is a slow burn. SPAC timelines don't match FDA timelines.
Nuro
AI Robotics† 2024$2.1B raisedWhat Happened
Autonomous delivery robot company faced massive layoffs and strategic pivot after burning through cash.
Lesson Learned
Last-mile delivery robots need unit economics that work, not just cool demos.
Cogitai
Foundation Models† 2018What Happened
Continual learning AI startup acquired by Sony. Focused on AI that learns from ongoing experience.
Lesson Learned
Continual learning research found a home in gaming and entertainment at Sony.
Woebot Health
AI Healthcare† 2024$114M raisedWhat Happened
AI therapy chatbot struggled with clinical validation and FDA regulatory pathway.
Lesson Learned
Mental health AI faces an impossible trust gap. Therapists don't trust bots, patients need trust.
Tractable
Computer Vision† 2024$115M raisedWhat Happened
AI for auto insurance claims. Struggled to expand beyond initial insurance use case.
Lesson Learned
Vertical AI needs to expand or die. One customer segment isn't a business.
Drive.ai
AI Robotics† 2019$77M raisedWhat Happened
Self-driving shuttle startup acquired by Apple in a fire sale after running out of cash.
Lesson Learned
Even Stanford AI PhDs couldn't solve autonomous driving on a startup budget.
Zoox
AI Robotics† 2020$955M raisedWhat Happened
Autonomous vehicle startup acquired by Amazon for $1.2B — a fraction of its $3.2B peak valuation.
Lesson Learned
Autonomous driving startups become acqui-hires. Amazon bought potential, not product.
Osaro
AI Robotics† 2023$28M raisedWhat Happened
AI-powered robotic picking for warehouses. Struggled to scale beyond pilot customers.
Lesson Learned
Warehouse robotics needs reliability at scale. Pilots don't equal contracts.
Leap Motion
Computer Vision† 2019$94M raisedWhat Happened
Hand tracking AI company sold to UltraHaptics for fraction of value. Consumer VR never took off as expected.
Lesson Learned
Input devices need killer apps. Amazing tech without a platform is a science project.
Afiniti
Enterprise AI† 2023$300M raisedWhat Happened
AI-powered call routing. CEO controversy and questionable AI claims led to collapse.
Lesson Learned
AI companies built on hype and personality cults don't survive scrutiny.
DataRobot
ML Platform† 2024$1B raisedWhat Happened
AutoML platform faced massive layoffs and valuation cuts. Enterprise AI sales cycles proved brutal.
Lesson Learned
Raising $1B creates expectations that enterprise AI sales can't meet. Growth at all costs backfires.
Gong.io
Enterprise AI† 2024$584M raisedWhat Happened
Revenue intelligence platform saw growth stall. Market became crowded with cheaper alternatives.
Lesson Learned
Conversation AI for sales is a feature, not a company. CRMs added it natively.
Turi
ML Platform† 2016$6.6M raisedWhat Happened
Machine learning platform (formerly GraphLab) acquired by Apple for ~$200M.
Lesson Learned
ML tooling companies are acquisition targets. Apple wanted the talent and IP.
Nara Logics
Enterprise AI† 2023$8M raisedWhat Happened
Nara Logics built a biologically-inspired recommendation engine but struggled to compete with simpler, cheaper recommendation alternatives from established platforms and cloud providers that offered good-enough solutions at lower cost.
Lesson Learned
Over-engineered AI that delivers marginal improvement over simpler alternatives loses in the market. Enterprise buyers prioritize ease of integration and total cost of ownership over algorithmic sophistication.
Kindred AI
AI Robotics† 2020$61M raisedWhat Happened
Kindred AI developed robotic manipulation systems using reinforcement learning for warehouse automation. After burning through significant capital, the company was acquired by Ocado Group in 2023 at a fraction of its peak valuation.
Lesson Learned
Robotic manipulation startups face extended development timelines before achieving commercial reliability. The gap between impressive demos and production-grade warehouse automation proved longer and more expensive than investors anticipated.
Body Labs
Computer Vision† 2017$10M raisedWhat Happened
Body Labs developed 3D body scanning and measurement technology using computer vision but struggled as a standalone business. Amazon acquired the company in 2017 for its body modeling technology to integrate into retail fitting solutions.
Lesson Learned
Highly specialized computer vision companies that solve one narrow problem often become features acquired by platform companies rather than sustainable standalone products generating their own recurring revenue.
Vicarious
AI Robotics† 2022$250M raisedWhat Happened
Vicarious spent over a decade pursuing general-purpose robotic manipulation through neuroscience-inspired AI but failed to achieve commercial viability. Alphabet acquired the company in 2022, absorbing the team into Google DeepMind.
Lesson Learned
Ambitious moonshot AI robotics ventures require patient capital and clear commercial milestones. A decade of research without revenue makes standalone operation unsustainable, even with world-class talent and deep-pocketed investors like Peter Thiel and Elon Musk.
Chorus.ai
Enterprise AI† 2021$100M raisedWhat Happened
Chorus.ai built a conversation intelligence platform for sales teams that analyzed calls to improve performance. ZoomInfo acquired the company for $575M in 2021, integrating its AI capabilities into a broader B2B intelligence platform.
Lesson Learned
Sales conversation intelligence proved more valuable as a feature within a larger data platform than as a standalone product. ZoomInfo $575M acquisition validated the technology while confirming the feature-not-product dynamic.
Dextro
Computer Vision† 2018$5M raisedWhat Happened
Dextro built computer vision technology for automated video analysis but could not find a sustainable enterprise market willing to pay for standalone video understanding capabilities in a pre-GPT-4V era.
Lesson Learned
Video understanding AI was a technically impressive capability that arrived before enterprise demand materialized. The lesson is timing — capabilities need matching market readiness to build a viable business.
Ginger.io
AI Healthcare† 2021$220M raisedWhat Happened
Ginger.io built AI-powered mental health monitoring using smartphone sensor data but struggled with healthcare monetization. The company merged with Headspace in 2021, diluting its original AI-first clinical vision into a broader wellness platform.
Lesson Learned
Mental health AI monetization proved significantly harder than expected due to regulatory complexity, insurance reimbursement challenges, and consumer willingness to pay. Pure AI plays in healthcare need clear payer strategies from day one.
Digital Genius
AI Agents† 2023$15M raisedWhat Happened
Digital Genius provided AI-powered customer service automation for enterprises but saw its market position collapse when ChatGPT commoditized conversational AI capabilities in late 2022, making its specialized solution less differentiated.
Lesson Learned
When OpenAI released ChatGPT, it disrupted an entire category of specialized conversational AI point solutions overnight. Vertical AI companies must build defensible moats beyond the base language model capability.
Determined AI
ML Platform† 2021$13.6M raisedWhat Happened
Determined AI built an open-source deep learning training platform that streamlined distributed training and hyperparameter tuning. Hewlett Packard Enterprise acquired the company in 2021 before it achieved meaningful independent commercial scale.
Lesson Learned
Open-source ML infrastructure companies face monetization challenges when cloud providers offer competing managed services. HPE acquisition provided an exit but confirmed the difficulty of building standalone ML platform businesses.
Distilled from every failure
Universal Lessons
Product-market fit beats fundraising
Companies that prioritized raising over building consistently ran out of time. Revenue is the only true runway extension.
Speed of execution is non-negotiable
In AI, a 6-month delay is an eternity. The companies that moved fastest — even imperfectly — outlasted those waiting for perfection.
Niche beats general
Broad 'AI for everything' companies collapsed under their own scope. The survivors went deep in one vertical and owned it.
Moat over model
Access to frontier models is now a commodity. Distribution, data, and distribution won. Model quality alone was never enough.
Common Questions
FAQ
What is the AI Graveyard?
Why do AI companies fail?
How is this data collected?
Can I submit a company to the graveyard?
Know a failed AI company?
Help build the most complete record of AI industry failures. Every post-mortem helps the next generation succeed.
Submit a Company