AI & Machine Learning · TruthScan + Claude
TruthScan MCP with Claude: Verify Media and Documents Before You Act on Them
The short answer
Claude works with TruthScan through the Model Context Protocol (MCP), an open standard that lets Claude use outside software securely. Once connected with an API key, Claude can ask TruthScan to check text, images, PDFs, video or audio for signs of AI generation or manipulation, then explain the result in plain language for your team to review.
What is TruthScan?
TruthScan is a content authenticity platform. It helps teams check whether images, text, PDFs, video and audio were created or altered by AI before they trust, publish or act on them. Typical users include editorial teams, claims and compliance staff, HR and recruitment teams, legal departments, and communications leaders who receive large volumes of material from outside sources and need a reliable first check.
How Claude connects to TruthScan through MCP
The Model Context Protocol is an open standard that acts as a universal adapter between Claude and other software. It lets Claude securely use a tool by reading information and taking actions, with permissions your organisation controls. For TruthScan, the connection is set up with an API key, so access is tied to a credential your IT team issues and can revoke at any time.
Once connected, Claude can send files or text to TruthScan and ask for checks in everyday language. For example, a staff member can ask Claude to review a batch of supplier invoices or a set of submitted photos, and Claude will run the relevant checks and summarise which items need a closer look. Claude can also compare results across several items and draft a short report.
Permissions and approvals keep people in charge. Access can be limited to the checks a given team needs. Claude presents findings and suggested next steps, but consequential decisions such as rejecting a claim, withdrawing a publication or escalating a case are routed to a named person for approval. Every request is logged so the organisation can see who asked for what and when.
What you can automate
Editorial image and text checks
Newsrooms and content teams receive images and written submissions from freelancers and contributors. Claude can pass each item to TruthScan before it reaches an editor, flag likely AI-generated or altered images, and note which text looks machine-written. Editors keep the final call on what gets published, and the audit log records each check.
Insurance claim document review
Claims teams often receive photos of damage and supporting PDFs. Claude can run TruthScan checks on these files as they arrive and flag any that show signs of manipulation. Adjusters then review the flagged items with the context Claude provides, which can shorten the first-pass review without removing human judgement from the claim decision.
Recruitment and HR document screening
Recruiters sometimes receive certificates, references and writing samples that need a basic authenticity check. Claude can submit these documents to TruthScan and summarise the results for the hiring manager. The check supports, rather than replaces, normal verification with issuing institutions and referees.
Legal and compliance evidence intake
Legal teams collecting evidence from employees, customers or third parties can use Claude to route audio, video and PDF submissions through TruthScan. The system highlights items that warrant forensic follow-up. Counsel decides how the findings are used, and the logged trail helps show how each file was handled.
Communications and brand protection
Communications teams can check images and audio that appear to come from their organisation, such as executive statements or product visuals, before they are shared internally or with the press. Claude can prepare a short summary of concerns so the team can respond quickly and calmly.
An end-to-end workflow: Supplier invoice and photo verification
TriggerA supplier e-mails invoices and proof-of-delivery photos to the accounts payable mailbox.
Claude collects the attachments and sorts them by supplier and purchase order number.
Claude sends each PDF and photo to TruthScan through the API key connection and requests authenticity checks.
TruthScan returns results, which Claude compares against the expected format and amounts in the ERP system.
Claude writes a summary listing clean items and any that need a closer look, with the reason for each flag.
An accounts payable officer reviews flagged items in Microsoft 365 and approves or rejects payment.
The decision and the check results are recorded in the ERP and in the audit log.
OutcomePayment teams spend less time on routine first-pass checks and focus their attention on the invoices that most need human review. Typical results depend on invoice volume and the quality of the source files.
How we set it up
Free automation audit
We review your current intake of images, documents, audio and video, and identify the checks that would help most. You receive a short written plan with suggested workflows and priorities at no cost.
Access and API key setup
Your IT team issues a TruthScan API key with the narrowest permissions that the agreed workflows need. We connect Claude to TruthScan through MCP, using the API key as the sign-in method, and store the credential securely.
Workflow design
We design the end-to-end process across TruthScan, Claude and your systems such as Microsoft 365, your CRM or ERP, and your ticketing tool. We define which decisions are automatic, which need approval, and who gives that approval.
Testing with real cases
We run the workflow against past cases your team has already resolved, compare outcomes, and tune thresholds and summaries. We agree results with the business owner before anything goes live.
Launch, monitoring and training
We launch with logging and monitoring turned on, watch early results closely, and train your staff on reviewing flagged items and reading Claude's summaries. Support continues as volumes and rules change.
Who benefits
- Editors and content leads
- A consistent first check on incoming images and text, so editorial judgement goes to the items that need it most.
- Claims and compliance managers
- Faster triage of submitted documents and media, with a clear audit trail for each decision.
- HR and recruitment heads
- A basic authenticity screen on certificates and documents, handled consistently across hiring teams.
- IT and security leaders
- Controlled access through a single API key, least-privilege permissions and full logging of activity.
Security and control
- Access uses an API key with the narrowest permissions needed for each workflow, and can be revoked at any time.
- People approve any consequential decision, such as rejecting a claim or withdrawing content.
- Every request and result is logged so the organisation can review activity.
- Detection results are treated as indicators for review, not as proof, and workflows are tested before launch to understand their error rates.
- Data handling rules are agreed in advance so sensitive files are only sent where your policies allow.
For teams in the US, UK, India and the Middle East
Organisations in the United States and the United Kingdom are subject to their own data protection frameworks, and we configure workflows to fit them. Indian organisations must meet the Digital Personal Data Protection Act, 2023, and businesses in the UAE and Saudi Arabia must follow their local data protection laws. We choose the hosting region for any custom application to meet these requirements, whether in India, the United States, the United Kingdom or the European Union.
Learn more about Claude and MCP
Anthropic's AI model and assistant.
Secure links between Claude and business tools.
How engineers build and change software with Claude.
Hosted agents that run multi-step work.
The open standard Claude uses to connect to tools.
Frequently asked questions
What does the TruthScan MCP server do?
The TruthScan MCP server lets Claude send content to TruthScan and receive authenticity results in a structured way. In general terms, it can support checks on AI-written text, images, PDFs, video and audio, and Claude turns those results into plain-language summaries. The exact checks available depend on how TruthScan is configured for your account.
Is connecting TruthScan to Claude secure?
The connection is designed around least-privilege access, a single API key tied to a defined set of permissions, and logging of activity. Your IT team controls and can revoke the key. We also agree data handling rules in advance so sensitive files are only processed where your policies allow.
What can Claude do with TruthScan once connected?
Claude can submit text, images, PDFs, video and audio for checking, interpret the results, compare items, and draft summaries or reports. It can also prepare the next steps for a person to approve. Claude does not make final decisions on consequential matters without a person signing off.
How long does setup take?
A typical project runs from the free audit through to launch and training over several weeks, depending on how many workflows are involved and how much testing against past cases is needed. Simple single-workflow connections can move faster. We confirm the timeline during the audit.
Which TruthScan plan or permissions do I need?
You need an active TruthScan account that can issue an API key, and permissions that cover the checks your workflows require. The right level depends on your volumes and the file types you handle. Your IT team and our audit will confirm what is needed before anything is connected.