Only a small subset of the retrieved evidence supports the exact request: a comparison of the most recent filing vs. the immediately prior comparable filing with quoted before/after language. Most batches contained only latest-filing snippets, so those names cannot be included without unsupported inference.
Below are the companies where the provided matches contain at least some explicit before/after disclosure language relevant to the requested themes.
| ticker | Company Name | Specific risk factor | Change type | Before language | After language | Source | Market Cap |
|---|---|---|---|---|---|---|---|
| TXG | 10x Genomics, Inc. | Single-source / sole-source supplier dependence | Reworded / materially softened | “We and our customers are dependent on single source and sole source suppliers for some of the equipment, components and materials used in our products and in conjunction with our products and the loss of any of these suppliers could harm our business.” | “We obtain some components of our instruments and consumables from third-party suppliers. While some of these components are sourced from a single supplier, we have qualified second sources for some, but not all, of our components including critical reagents, enzymes and oligonucleotides.” | <a href="https://www.sec.gov/Archives/edgar/data/1770787/000162828026007822/txg-20251231.htm">2026-02-12 10-K</a> | $5.88B |
| AJG | Arthur J. Gallagher & Co. | AI-related risk / AI competition / AI-enabled cyber threats | Reworded / materially expanded | “We are subject to risks associated with AI.” / “These risks include, among others, the input of confidential information… in violation of applicable laws or regulations, including those relating to data protection and AI.” | “The increasing adoption of AI technologies by cyber threat actors presents a significant and evolving risk to our company. These actors may leverage AI to develop more sophisticated and targeted cyberattacks…” / “AI and its applications are developing rapidly… the use of this technology by our clients or underwriting enterprises may impact the way our business operates, and its use by our competitors and new market entrants… may give them a competitive advantage.” | <a href="https://platform.hudson-labs.com/company/354190/transcript/901727/full">2026-02-17 10-K</a> | $69.09B |
| SHC | Sotera Health Company | Limited counterparty / supplier dependence | Reworded / expanded | “We depend on a limited number of counterparties to provide the materials and resources we need to operate our business.” | “We depend on a limited number of counterparties to provide the materials and resources we need to operate our business. Any disruption in the availability of, or increases in the price of, EO, Co-60 or our other direct materials, services and supplies, including as a result of geopolitical instability or sanctions against Russia by the United States, Canada, the United Kingdom and/or the European Union, or sanctions by Russia against those countries, may have a material adverse effect on our operating results.” | <a href="https://www.sec.gov/Archives/edgar/data/1822479/000182247926000015/shc-20251231.htm">2026-02-24 10-K</a> | $5.00B |
| NKE | NIKE, Inc. | AI-related data security / privacy risk | Reworded / updated | 2025: “We are subject to data security and privacy risks that could negatively affect our results, operations or reputation.” / “The risks of inadvertent disclosure of personal data can increase with the introduction of new and complex technologies, such as artificial intelligence features…” | 2026: “We are subject to data security and privacy risks that could negatively affect our results, operations or reputation.” / “Regulations and best practices with respect to new technological developments, including generative AI, are in the process of being developed globally.” / “Threat actors are increasingly sophisticated…” | <a href="https://platform.hudson-labs.com/company/320187/transcript/949980/full">2026-07-15 10-K</a> | $64.12B |
| NP | Neptune Insurance Holdings Inc. | Capacity provider concentration / dependence | Reworded with mitigation language; arguably softened in tone | “Our business may be harmed if one or more of our relationships with capacity providers are terminated or are reduced, if we fail to maintain good relationships with such capacity providers, if we become dependent upon a limited number of capacity providers, or if we fail to develop new capacity provider relationships.” | “Our business may be harmed if one or more of our relationships with capacity providers are terminated or are reduced, if we fail to maintain good relationships with such capacity providers, if we become dependent upon a limited number of capacity providers, or if we fail to develop new capacity provider relationships.” The filing also adds: “We mitigate this risk by diversifying our capacity provider panel…” and “our insurance agent and broker partners drove over 96% of our policy sales…” | <a href="https://platform.hudson-labs.com/company/2067129/transcript/904820/full">2026-02-25 10-K</a> | $4.82B |
| GEV | GE Vernova Inc. | Capacity-related competition risk in power equipment markets | Reworded with mitigation language; arguably softened in tone | “Due to increasing demand exceeding available capacity for products and services that supply the electrical power industry, we face growing competition from emerging threats.” | “Due to increasing demand exceeding available capacity for products and services that supply the electrical power industry, we face growing competition from emerging threats.” The filing also says: “The continuing ability to reduce cycle times and ensure available capacity is expected to allow us to remain competitive…” | <a href="https://www.sec.gov/Archives/edgar/data/1996810/000199681026000015/gev-20251231.htm">2026-01-29 10-K</a> | $239.77B |
Best-supported “quiet changes”
- 10x Genomics (TXG) is the clearest example of a material softening in a requested category. The prior filing said the company and customers were “dependent on single source and sole source suppliers,” while the new filing reframes this to say only “some” components are single-sourced and adds that it has “qualified second sources for some, but not all” components.
- Arthur J. Gallagher (AJG) shows a clear AI risk-factor expansion from generic AI risk language to more specific disclosure about AI-enabled cyberattacks and competitors/new entrants gaining advantage from AI.
- Sotera Health (SHC) keeps the limited-counterparty concept but expands it with more concrete supplier/material exposure, including EO, Co-60, and sanctions-related disruption.
- Nike (NKE) shows an updated AI/privacy/cyber framing: the older wording emphasized inadvertent disclosure risk from AI features, while the newer wording shifts toward generative AI regulation and increasingly sophisticated threat actors.
- Neptune Insurance (NP) and GE Vernova (GEV) are notable because the retrieved evidence suggests the core risk remained, but the newer filing added mitigation-oriented language, which can read as a subtle softening in tone even though the underlying risk statement was retained.
What the retrieved evidence does not support
- No company in the supplied matches has a clearly evidenced removed risk factor with quoted prior language and proof of omission in the latest filing.
- Many companies in the batches appear to have new AI, customer concentration, supplier dependence, or data-center-capacity language, but the matches usually lack the prior filing text, so they cannot be reliably classified here as added/removed/reworded with before/after quotes.
Partial-evidence names that look relevant but are not fully supportable here
The batches surfaced many likely candidates—such as UAL, WDC, SMTC, AMT, DLR, CLS, CRWV, GLXY, NCNO, BLCO, GEHC, SailPoint, SharkNinja, Rubrik, Waystar, and others—for single supplier dependence, customer concentration, AI competition, or data-center capacity commitments. But the provided snippets generally include only the latest filing language, not the prior comparable filing text needed for a compliant before/after comparison.
If useful, the next step would be a stricter shortlist of only the strongest “materially softened” changes, which from the supplied evidence would mainly be TXG, NP, and GEV.
