A request for verification and supporting evidence.

Dear Marici Team,

My name is published under a pseudonym for source-protection reasons. I am an independent investigative journalist conducting a review into nonprofit organizations making large-scale claims relating to AI-driven anti-trafficking operations, predictive analytics, behavioral intervention technologies, and donor-funded intelligence systems.

I am contacting you regarding statements and representations published on your website, promotional materials, donor messaging, and other public-facing communications, including but not limited to:

• “The world's most sophisticated AI”
• “50 AI tools,” including an AI Intelligence Analyst, AI Prosecutor, and AI Behavioral Scientist
• Predictive mapping described as operating at 80%+ accuracy
• A reported 90%+ reduction or “collapse” of child sex trafficking across six unnamed cities
• Thousands of trafficking networks reportedly neutralised
• AI-powered rehabilitation and aftercare systems
• A stated partnership with Stanford University's “Precision Medicine Department” on a landmark longitudinal study using one million data points
• Claims of intelligence-led operational intervention and behavioural prediction systems
• Donor-page representations equating specific dollar amounts to specific numbers of lives saved

Given the scale and seriousness of these claims, and the fact they are being presented to donors, supporters, and the general public, I would appreciate clarification and supporting documentation regarding the following areas.

1. Technology Stack, Infrastructure, and AI Systems

Please clarify:

• Whether your systems are proprietary, licensed, outsourced, or built using third-party platforms
• Which AI models, APIs, cloud providers, or software vendors are currently used
• Whether your infrastructure relies on providers such as OpenAI, Anthropic, Google, AWS, Azure, Palantir, ArcGIS, Salesforce, Databricks, Snowflake, Airtable, Supabase, or similar services
• Whether any systems use large language models, predictive analytics engines, facial recognition, NLP pipelines, sentiment analysis, graph databases, or behavioural classification systems
• Whether your “50 AI tools” are independently developed or wrappers / integrations around third-party APIs
• Whether any components are commercially licensed products rather than internally developed technology
• Whether any patents, proprietary algorithms, or model architectures exist
• Whether any code repositories, technical papers, whitepapers, or architecture overviews are available for review
• Whether your systems have undergone external cybersecurity review or penetration testing
• Whether your systems process real-time law-enforcement or victim-related data
• What data governance, consent, privacy, retention, and safeguarding frameworks are in place

2. Evidence Supporting Impact Claims

Please provide evidence supporting public claims including:

• “90% collapse” or similar reductions in trafficking activity
• “Thousands of networks neutralised”
• “Operational success” across multiple cities
• Predictive-mapping accuracy claims
• Behavioural prediction or intervention success rates

Please include:

• Independent audits or evaluations
• Statistical methodology and baseline measurements
• Names of operational jurisdictions or cities
• Date ranges used in calculations
• Third-party validation or oversight
• Peer-reviewed studies or academic publications
• Outcome metrics and longitudinal tracking data
• False-positive and false-negative rates where applicable
• Documentation demonstrating causal links between your interventions and claimed reductions

3. Law Enforcement, Government, and Institutional Relationships

Please clarify:

• Which law-enforcement agencies, NGOs, prosecutors, or institutions currently use or have tested your systems
• Whether any memoranda of understanding (MOUs), pilot agreements, or operational partnerships exist
• Whether your technology has resulted in arrests, prosecutions, convictions, or measurable safeguarding outcomes
• Whether any agencies have independently validated your systems
• Whether any government grants or contracts have been awarded
• Whether any external ethics or oversight committees review operational activity

4. Stanford Partnership Claim

Your Aftercare page appears to state that Marici has partnered with Stanford University's Precision Medicine Department on a landmark longitudinal study. Please provide verification of this partnership, including:

• The specific Stanford department, school, institute, or programme involved (Stanford does not appear to operate a unit literally named “Precision Medicine Department”)
• The named principal investigator and their Stanford affiliation
• IRB approval number or study registration
• Scope of work and publication timeline
• Whether Stanford formally endorses or independently validates Marici's aftercare model or impact claims
• Any Stanford-issued communication confirming the partnership

5. Financial Transparency and Governance

Please clarify:

• Whether two separate EINs (82-1536804 and 85-3455940) were at any time in circulation on Marici donor-facing materials, and if so, the classification of donations made under the unmatched EIN
• All legal entities operating under the Marici name (including any foreign-registered affiliates such as the Cambodian FNGO registration)
• Whether all donor funds are directed to a single registered nonprofit entity
• Availability of audited financial statements and the name of the auditing firm
• Board oversight and governance structure, including a public list of board members beyond the four named officers
• Whether any independent audit committee exists
• Executive compensation structure
• Vendor relationships and major technology expenditures
• Allocation percentages between technology development, fundraising, administration, and field operations
• The cause of the FY2024 operating loss of approximately $1.33 million

6. Operational Demonstrability

Please provide:

• Screenshots, demonstrations, dashboards, or technical walkthroughs of the systems being referenced publicly
• Evidence that the AI systems described are operational rather than conceptual or aspirational
• Case studies with independently verifiable outcomes
• Documentation showing how frontline intervention decisions are generated or supported by your systems

7. Donor Transparency and Public Accountability

As these claims are being used in donor-facing communications, please also provide:

• Copies of annual reports
• IRS Form 990 filings (currently available only through third-party databases)
• Independent impact assessments
• Safeguarding policies
• Data-protection policies
• Survivor outcome metrics
• Internal ethics frameworks governing AI use
• Any third-party reviews, complaints, or investigations relating to your claims or operations

8. Donor Page Impact Representations

I want to ask specifically about the impact representations made on your donation page, which states “Your gift ends an industry” and “Every dollar saves children and transforms cities,” and which presents specific giving levels using direct life-saving language (“Save 1 Life — $500” through to “Save 50 Lives — $25,000”).

This appears to create a direct donor-facing representation that $500 saves one life and that higher donations save a proportionate number. Please provide the evidence, methodology, and internal calculation used to support this claim. Specifically:

• How is “Save 1 Life” defined?
• Does $500 directly fund a specific rescue, intervention, prosecution, aftercare placement, or verified safeguarding outcome?
• Is the $500 figure based on audited historical cost-per-outcome data?
• Who independently verifies these life-saving outcomes?
• How many individuals were actually “saved” in the most recent financial year?
• How many of those outcomes can be tied directly to donor-funded activity?
• What percentage of each $500 donation goes to direct intervention versus administration, fundraising, technology, salaries, consultants, travel, marketing, or overhead?
• Are donors informed that the “Save X Lives” language is an estimate, projection, symbolic figure, or verified operational outcome?
• Has this donation language been reviewed by legal counsel, nonprofit fundraising compliance specialists, or any state charity regulator?

If these claims are symbolic or aspirational rather than evidence-based, please clarify why the donation page does not clearly say so.

This inquiry is part of a broader review into transparency, accountability, and evidence standards surrounding AI-driven nonprofit claims in the anti-trafficking sector.

For accuracy and fairness, I am happy to publish your full response, clarifications, and supporting evidence in this review, unedited and in full, regardless of length.

I would appreciate a response within seven days.

Kind regards,