What public companies just told the SEC about AI, read straight from the filings. Every quote links the original document.
Anthropic Files S-1 at $70 to $78 Per Share
The most closely watched AI IPO in years moved from private to public record on November 10. Anthropic, PBC filed an amended S-1 setting a preliminary price range of $70.00 to $78.00 per share and a planned Nasdaq listing under the ticker "ANT." The public benefit corporation structure is not window dressing; it appears in the S-1 as a legal obligation with real consequences for how the board must weigh shareholder returns against mission.
We are incorporated as a public benefit corporation under Delaware law, a structure consistent with our commitment to artificial intelligence ("AI") safety and responsibility. As a public benefit corporation, we are required to balance the pecuniary interests of our stockholders, the best interests of those materially affected by our conduct, and the specific public benefit purpose identified in our amended and restated certificate of incorporation: the responsible development and maint...
That statutory balancing test will be a live governance question for investors once shares trade. PBC status means boards can cite the public benefit purpose when resisting shareholder pressure, but it also means plaintiffs can sue if the company demonstrably abandons it.
Agentic AI Named as Direct Threat to Existing Revenue
Two companies used this week's filings to flag agentic AI specifically as a force that could hollow out their core businesses, language that goes beyond standard risk-factor boilerplate.
FiscalNote Holdings, which sells policy-intelligence subscriptions to governments and enterprises, disclosed in an 8-K:
the risk that general purpose generative AI platforms and agentic AI tools will directly compete with and reduce demand for custom-built SaaS tools and subscription products
FiscalNote is acknowledging in a public filing that the category it occupies may be absorbed by general-purpose AI agents. That is a meaningful disclosure for a company whose recurring revenue depends on customers staying subscribed to purpose-built tools rather than prompting a foundation model instead.
Agricultural chemicals company American Vanguard named the same technology from an entirely different angle in its 10-Q. The company does not sell AI products; it runs internal systems like most industrials. What it disclosed was a specific asymmetry argument:
The development of Agentic AI tools may overtake the efficacy of available cybersecurity defense tools and, as such, could put the Company's computing systems at risk. Fueled by enormous investment and activity from competing global developers, agentic AI solutions are evolving at a rapid pace and becoming increasingly powerful. By contrast, cybersecurity defense tools are evo...
An agricultural inputs company reasoning in a quarterly filing about offensive AI outpacing defensive tooling is worth flagging. The underlying concern is real across sectors; American Vanguard is among the earlier industrial filers to put it on record with this level of specificity.
Agentic AI as a Product Bet at Trade Desk and SoundHound
The Trade Desk CEO Jeff Green named "agentic AI" in earnings remarks filed via 8-K, placing it alongside specific platform features Zuma and Audience Unlimited as improvements intended to restore the company's growth trajectory. The Trade Desk reported a weaker quarter and Green is explaining what will drive recovery. Naming a specific technology bet in an 8-K earnings release is a form of forward guidance investors will hold the company to.
SoundHound AI disclosed in its 10-Q that last year's acquisition of Interactions Corporation was intended to "strengthen and extend SoundHound's growing leadership in Agentic AI," describing Interactions as "a pioneer in AI for customer service and workflow orchestration." For a company whose valuation is closely tied to AI sentiment, articulating the acquisition thesis around agentic AI leadership creates a measurable bar for future performance disclosures.
Tonix Pharmaceuticals Puts AI Drug-Development Risk on the Record
Tonix Pharmaceuticals included a named risk section in its 10-Q rather than a passing reference in a list. The language addresses what can go wrong in a regulated development pipeline:
Our use of artificial intelligence could adversely affect our business, financial condition and results of operations We use artificial intelligence ("AI") and computational tools in research, drug development and corporate activities. Inaccurate or biased outputs may remain undetected despite human review, misdirect resources or delay development and regulatory submissions.
For a clinical-stage company, a delayed regulatory submission is a material financial event, not an abstract technology risk. Seeing this specificity from a small-cap pharma rather than a software company reflects how broadly AI is now embedded in regulated workflows.
PowerCompute Reports Its First Month Under a Rebranded AI Mission
PowerCompute, formerly LM Funding America, filed its first monthly operations report under its new identity as a company developing HPC and AI infrastructure alongside Bitcoin mining. For October 2026 it mined 9.0 Bitcoin (up 20% year-over-year), held 65.7 Bitcoin on its balance sheet, and generated approximately $46,000 from energy sales. The AI infrastructure ambitions appear in the company name and the filing's description of the business; October's operating metrics are still pure Bitcoin mining. The company trades on Nasdaq as PWCM.