SEC 10-K risk factor changes with AI

Gather the evidence

Read the same section in both original filings and align by topic rather than heading order. Classify new topics, expanded or narrowed wording, removed passages and unchanged topics. Missing text from a partial extract is not proof that a disclosure was removed.

Use Search and Contents if your application already handles the analysis. To generate the brief and CSV together, run the DeepResearch request below.

Try it

Set VALYU_API_KEY in your environment. This is an illustrative request, not a saved result. Replace the company, periods or sources for your own task.

Run the research request

cURL

curl --fail-with-body https://api.valyu.ai/v1/deepresearch/tasks \
  -H "x-api-key: $VALYU_API_KEY" \
  -H "Content-Type: application/json" \
  --data-binary @- <<'JSON'
{
  "query": "Compare Item 1A Risk Factors in Microsoft FY2024 and FY2025 Forms 10-K. Align matching topics despite heading or order changes. Identify new topics, expanded or narrowed wording, removed passages and unchanged topics. For each claimed change provide short before and after passages and both filing URLs or section references. Do not infer that a described risk occurred. Mark incomplete extraction rather than treating missing text as deletion.",
  "mode": "standard",
  "research_strategy": "Use the two original SEC annual filings and full relevant risk-factor sections where retrievable. Verify filing year, entity and section completeness before comparing. Check whether a passage moved or is incorporated by reference before labeling it removed.",
  "report_format": "A concise evidence brief and comparison table. Cite each material finding, retain missing information and separate direct disclosure from interpretation.",
  "deliverables": [
    {
      "type": "csv",
      "description": "Populate the requested evidence fields with source references. Preserve unknown values and do not invent missing disclosures.",
      "columns": [
        "risk_topic",
        "change_type",
        "earlier_fiscal_year",
        "later_fiscal_year",
        "earlier_passage",
        "later_passage",
        "earlier_source_url",
        "later_source_url",
        "section_reference",
        "review_note"
      ]
    }
  ]
}
JSON

Save the returned deepresearch_id and poll task status. When status is completed, read output and download the CSV from deliverables; handle failed tasks separately.

Check before using

The filings, fiscal years and section boundaries are correct.

Reordering, formatting and extraction gaps are not labeled substantive changes.

Text changes are kept separate from claims that an incident occurred or a risk became more likely.

Keep the accepted output and its evidence together in your analyst workspace. See Research in your application for integration details.

Start here

APIs

Finance workflows

Sciences workflows

Build and connect

Comparisons

SEC risk-factor changes

Example

Gather the evidence

Try it

Check before using

​Example

​Gather the evidence

​Try it

​Check before using

Example

Gather the evidence

Try it

Check before using