The Tool Desk
Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
Yes, some AI-data firms have reportedly acquired code from failed startups for resale or packaging for AI labs, and buyers have shown interest in broader records of how companies worked. But the evidence does not establish a market-wide buying trend or show that every reported approach became a sale to an AI lab. The key distinction is between documented acquisitions, offers, and discussions.
What is reportedly being sold?
The reported market goes beyond source code. The Information has reported that data firms including Turing and AfterQuery acquired codebases from failed startups to resell or package for AI labs. Its reporting also describes interest in operational records such as Slack conversations, tickets, code, and other records of work. The Information’s articles are subscriber-restricted, so these claims should be understood as attributed reporting rather than a public transaction ledger.
A seller-facing marketplace page lists private repositories, datasets, tasksets, reinforcement-learning environments, and complete workspaces from closed companies. A workspace might include Slack, email, tickets, wikis, or finance records. Those are categories the commercial service says it markets; the page does not independently establish that every category has been sold. See the seller-facing asset categories.
| Asset | What it could contain | What the reporting establishes |
|---|---|---|
| Code repository | Source code and associated development context | The Information reported acquisitions of failed-startup codebases by data firms for resale or packaging to AI labs. |
| Operational records | Communications, tickets, work records, and workflows | The Information reported buyer interest; the available evidence does not establish that every approach resulted in a purchase. |
| Complete workspace | Potentially Slack, email, tickets, wikis, and finance records | A commercial seller page lists these categories, but its listing is not independent proof of completed sales. |
How strong is the evidence of actual purchases?
The reporting describes different stages that should not be collapsed into one claim. A reported acquisition by a data firm is not automatically a completed purchase by an AI lab. An offer or approach is not a negotiation, and a negotiation is not a signed deal. When a report describes interest or discussions, it is more accurate to say that a buyer was exploring a deal than that the buyer bought the data.
#1 Best Overall
- Reported acquisition: The Information reported that data firms acquired some failed-startup codebases for resale or packaging.
- Approach or offer: A firm may seek access to an archive without obtaining it.
- Discussion: Talks about a potential license or partnership do not prove that an agreement was completed.
- Completed lab purchase: The available sources do not establish that every archive acquired by an intermediary was ultimately bought by an AI lab.
No reliable aggregate count or total market value is established by the reviewed reporting. That is a limit of the available evidence, not proof that no sales are happening. Individual offers should not be treated as typical prices or used to estimate an industry-wide total.
Why might AI companies want company records?
Code can show how software was built, while tickets and communications can preserve the decisions, tasks, and workflows surrounding that code. A broader archive may therefore be marketed as more contextual than a repository alone. That is a potential distinction in the asset being offered, not evidence that a particular AI lab trained a model on any specific archive.
Rank #2
- Use scikit-learn to track an example ML project end to end
- Explore several models, including support vector machines, decision trees, random forests, and ensemble methods
- Exploit unsupervised learning techniques such as dimensionality reduction, clustering, and anomaly detection
- Dive into neural net architectures, including convolutional nets, recurrent nets, generative adversarial networks, autoencoders, diffusion models, and transformers
- Use TensorFlow and Keras to build and train neural nets for computer vision, natural language processing, generative models, and deep reinforcement learning
Completeness matters too: a repository without its history or surrounding work records is different from a workspace containing several systems and years of activity. The available sources describe categories of material and buyer interest, but do not establish how any specific buyer evaluates quality, what was used to train a model, or whether an acquired dataset was ultimately used for training.
How does this differ from data deals with active companies?
Separate reporting by The Information describes OpenAI and Anthropic discussing specialized data deals, including genomics, with companies. It also describes conversations involving Google DeepMind and biotechnology startups. These reports broaden the story beyond failed businesses, but they concern discussions—not necessarily completed purchases—and do not show that the participants acquired archives from defunct startups.
Rank #3
| Situation | What is being discussed | Evidence and caveat |
|---|---|---|
| Closed or failed startup | Code and potentially broader records of company operations | The Information reported codebase acquisitions by data firms and interest in operational records; a downstream lab purchase is not established for every case. |
| Active company | Specialized data, including genomics, potentially through a license or partnership | The Information reported discussions involving AI labs and companies; discussions alone do not establish completed deals. |
These arrangements can also differ in form. A direct sale transfers an asset under agreed terms, while a license or partnership may provide limited or conditional access. The available reporting does not give enough transaction-specific detail to characterize every arrangement as a sale, license, or completed partnership.
Who can authorize the sale, and what privacy questions arise?
A company shutting down does not, by itself, establish who owns or controls its code and records. The seller-facing service says a seller should own an asset outright or have written permission from contributors, and says employees cannot sell an employer’s code or data. Those statements describe that market participant’s stated screening practice; they are not a legal opinion or a universal rule.
Rank #4
Rights questions can involve more than the company’s code. An archive may include employee communications, customer information, contributor work, or records governed by contracts. The available sources do not resolve who controls any particular archive after insolvency or acquisition, what a company’s agreements allow, or what consent is required in a given jurisdiction. A buyer or seller would need to assess the specific ownership, permissions, contracts, and sensitivity of the material rather than assume closure clears those issues.
Crashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteWindows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallPrivacy is a separate concern from ownership. A company workspace can contain information about people who were not parties to a proposed data deal. The available coverage does not establish a universal answer for whether or how such records can be used for model training; the answer depends on the archive and applicable agreements and rules.
Best Value
How does this relate to consumer-data marketplaces?
On August 6, 2026, Axios reported that Verb launched a consumer-data marketplace where people can select activity categories and offer personal data to companies. Axios noted privacy concerns and reported that the marketplace excludes categories such as passwords, text messages, and health data from analysis. Verb is an adjacent example of attempts to commercialize data; it is not evidence that the company buys dead-startup archives.
Axios quoted Verb CEO and co-founder Cyrus Beschloss saying, “Verb is the place that people ‘will go to monetize something that for decades we’ve given away for free.’” That comment concerned Verb’s consumer-data marketplace, not purchases of defunct companies’ records. Read Axios’s report on Verb.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
Recommended Free Tools
