Anthropic has launched Claude for Manufacturing, a new industry-specific plugin that connects its AI assistant to software used by manufacturers to automate administrative and operational workflows. The product includes 17 manufacturing-focused skills and four connectors that allow Claude to work with information stored in Autodesk Fusion, Infor, Microsoft 365 and Oracle NetSuite. The launch marks Anthropic’s latest effort to bring its AI technology into specialised business environments beyond general-purpose chat and coding applications.
Announced on November 12, 2026, Claude for Manufacturing is designed to help factory teams handle tasks such as preparing customer quotations, investigating quality failures, comparing suppliers, diagnosing equipment breakdowns and compiling production reports. Instead of requiring employees to collect information manually from multiple systems, the plugin can retrieve relevant records, analyse them and prepare draft documents for review. Anthropic says employees remain responsible for approving important decisions, and the plugin does not directly operate physical factory equipment.
What Is Claude for Manufacturing?
Claude for Manufacturing is an industry-specific plugin built for manufacturers that want to use AI across existing business software and operational processes.
The plugin combines 17 specialised skills covering several areas of manufacturing, including:
- Sales and customer quotations
- Quality management and inspection reports
- Factory operations and production reporting
- Supply chain and supplier sourcing
- Product design and engineering
- Equipment maintenance and troubleshooting
These skills provide Claude with instructions for completing specific manufacturing tasks. Some handle individual assignments, while others combine multiple steps into a longer workflow.
For example, an estimator receiving a new customer request can ask Claude to examine the requirements, identify missing information, compare costs from previous jobs and prepare a draft quotation.
Similarly, a quality engineer investigating a failed inspection can use Claude to document the problem, search previous reports for similar defects, identify possible causes and prepare a corrective-action report.
The objective is to reduce repetitive administrative work while allowing experienced employees to concentrate on decisions that require technical knowledge and judgement.
Four Connectors Bring Claude Into Existing Factory Software
One of the most important parts of the launch is the integration with software already used by manufacturing companies.
Claude for Manufacturing includes four connectors:
| Software | Manufacturing use case |
|---|---|
| Autodesk Fusion | Design comparisons and engineering variations |
| Infor | Purchase orders, inventory and production information |
| Microsoft 365 | Procedures, emails, reports and shared documents |
| Oracle NetSuite | Job costs, inventory, orders and production records |
These integrations allow Claude to work with information already maintained by manufacturers instead of requiring companies to move their operations into a new platform.
For example, an estimator using Oracle NetSuite can ask Claude to prepare a quotation based on actual inventory, previous orders, production steps and historical job costs.
An engineer working with Autodesk Fusion can request design variations based on an existing model and compare them against specified engineering targets.
The Microsoft 365 connector also allows Claude to retrieve relevant information from shared folders and email, bringing documents and records into a single workflow.
Manufacturers using other software can also work with Claude by saving regular data exports into Microsoft 365. The AI can then analyse those files even when a dedicated connector is unavailable.
Claude Can Help Diagnose Factory Equipment Problems
Equipment breakdowns can interrupt production, increase costs and require maintenance teams to investigate large amounts of technical information.
Anthropic says Claude for Manufacturing can help technicians organise that information when a machine stops working.
A technician can provide sensor exports, maintenance records, photographs and equipment manuals. Claude can examine the available material, rank possible causes and suggest what the technician should investigate first.
The model can also help production managers prepare reports showing output against targets, equipment downtime and the estimated financial impact of missed production.
These capabilities are particularly relevant for factories where maintenance teams spend considerable time gathering information from different systems before identifying the source of a problem.
However, Claude provides analysis and recommendations rather than independently repairing equipment or controlling machinery.
Infor Integration Can Help Manufacturers Respond to Supplier Delays
Anthropic’s manufacturing launch also includes an important connection with Infor, a provider of enterprise software used by manufacturers.
The standard Infor connector in Claude for Manufacturing allows teams to look up relevant records. Infor is also releasing a separate plugin that enables additional actions within its software.
Anthropic provided an example involving a delayed steel shipment.
If a supplier informs a manufacturer that a delivery will arrive two weeks late, a buyer can ask Claude to identify which customer orders may be affected and whether other facilities have available steel.
With the separate Infor plugin, the buyer can then approve actions such as releasing a purchase order to another supplier and rescheduling affected production.
The important distinction is that these actions require approval. Claude does not independently make purchasing or production decisions.
The integration also follows the buyer’s existing Infor permissions, helping companies maintain their established access controls.
Human Approval Remains Central to the System
Manufacturing is different from many other industries because mistakes can affect physical products, production schedules, equipment and worker safety.
Anthropic has therefore designed Claude for Manufacturing around human review and approval.
The plugin prepares drafts and recommendations, but employees remain responsible for checking the information before it reaches customers, suppliers or other business processes.
For example:
- An estimator reviews and releases a customer quotation.
- A quality engineer verifies a corrective-action report.
- A technician decides which suspected equipment problem to investigate.
- A buyer approves supplier-related changes.
Anthropic also says Claude can link figures and claims in its output to the underlying documents and records. This gives employees a way to check where information originated before acting on it.
The approach positions Claude as an assistant for manufacturing professionals rather than a replacement for the systems and people responsible for running a factory.
Anthropic Is Separately Developing Technology for Physical Equipment
Although Claude for Manufacturing connects AI to industrial software, it does not directly control machines.
Anthropic is separately developing something called the Model Hardware Standard, currently available as a research preview.
The proposed standard is intended to give programmable physical equipment a consistent way to describe its measurements, settings and safety limits to AI systems.
This could eventually help AI agents interact with devices such as robotic arms and laboratory equipment through standardised interfaces.
However, the Model Hardware Standard is separate from the manufacturing plugin. Anthropic has not announced that Claude for Manufacturing can independently operate production machinery.
The distinction is important because accessing factory records and controlling physical equipment involve very different technical and safety requirements.
Availability and How Manufacturers Can Get Started
Anthropic says Claude for Manufacturing is available to customers on Team and Enterprise plans.
Companies can install the plugin and ask Claude to help configure it. The assistant then asks which software systems the organisation uses, tests the available connectors and identifies skills that may require additional setup.
Manufacturers can adapt the supplied skills to match their own procedures, forms and operational requirements.
This flexibility is important because factories often have different production processes, quality standards and internal reporting requirements.
Anthropic also plans to add more manufacturing skills over time.
The Bigger Picture
Claude for Manufacturing reflects a growing shift in enterprise AI development. Instead of offering one general-purpose assistant for every industry, AI companies are increasingly building specialised products around the software, terminology and workflows used by particular sectors.
Manufacturing is a significant opportunity because employees frequently need to combine information from enterprise resource planning systems, engineering software, supplier communications and production records. Connecting these sources could reduce the time spent searching for information and preparing repetitive documents.
The launch also highlights an important distinction between AI assistance and industrial automation. Claude can prepare recommendations and coordinate software-based workflows, but human experts remain responsible for approving important actions. Anthropic’s separate Model Hardware Standard suggests that the company is exploring how AI could eventually interact more directly with physical devices.
Looking Ahead
Anthropic’s manufacturing plugin gives the company a more focused entry into industrial enterprise software, where customers may value practical improvements in quoting, quality management, maintenance and production reporting more than general-purpose AI features. Its success will depend on how reliably Claude works with real factory records, how easily companies can integrate their existing systems and whether the resulting time savings justify adoption costs.
The longer-term opportunity could extend beyond preparing reports and analysing records. If standards for connecting AI agents to physical equipment mature, AI systems may eventually participate in more complex industrial processes. For now, Claude for Manufacturing represents a software-focused step towards that possibility, with human approval and existing factory systems remaining central to how the product works.
Get the day’s top stories in your inbox
One concise email. No spam, unsubscribe anytime.



