What public companies just told the SEC about AI, read straight from the filings. Every quote links the original document.
Anthropic Files S-1 Defining Agentic AI for Public Markets
Anthropic, PBC filed its S-1 registration statement on October 26, entering the public filing record for the first time. The document includes a glossary that does something no regulator or standards body has managed to do cleanly: define the terms that have been spreading across corporate disclosures without consistent meaning. The company's own definition of the technology it builds:
"agentic AI" refers to AI systems that, with varying degrees of autonomy, can plan, determine, and sequence actions and execute them to accomplish defined objectives, including through interaction with external tools, services, datasets, and environments on behalf of a user.
The filing also defines "agent loop" as the iterative cycle in which an agent receives a task, takes an action, observes the result, and decides its next step. These definitions now exist in a federal securities filing. Regulators and courts looking for a baseline will find one here.
SAP Pays $1.26 Billion to Build the Agentic Data Layer
SAP's 6-K filed October 26 confirms the Reltio acquisition closed at $1,259 million (approximately €1,076 million), paid mostly in cash. SAP frames the deal in direct terms:
The acquisition is expected to help SAP to strengthen SAP Business Data Cloud (SAP BDC) and accelerate the evolution of SAP BDC to a fully interoperable enterprise data platform for enterprise-wide agentic AI and accelerate its customers' ability to govern and expose master data as trusted and context-rich data products across multiple sources that serve both traditional analytics workloads and AI agents.
SAP notes it is still in the process of identifying and measuring Reltio assets and liabilities, so purchase price allocation is pending. The acquisition positions master data governance as the foundational layer agents need to function reliably across enterprise systems.
"Agentic AI" Spreads Into Quarterly and Offering Filings
Omnicom's 10-Q filed October 23 is one of the more direct statements from a large incumbent. The advertising holding company, which reported over $15 billion in annual revenue, wrote into its quarterly filing:
Generative AI and agentic AI have, and we believe will continue to have, a significant impact on how we provide services to our clients and how we enhance the productivity of our people.
Workday's 424B5 prospectus supplement, a debt offering filed the same day, opens its company description with language investors can hold the company to:
Workday is the trusted agentic AI platform that operates at the core of human resources, finance, and information technology.
That is not a press release claim; it is in an offering document. RingCentral's 10-Q, also filed October 23, takes the opposite tone, flagging that it and its competitors "continue to invest significantly in AI (including machine learning, large language models, and agentic AI)" and acknowledging there is "no assurance that using AI in our platforms and products will enhance or be beneficial to our business." The candor in a risk factor is notable given how much investment RingCentral has publicly committed to the category.
An Insurance Broker Proposes to Enter GPU Cloud
Reliance Global Group's 8-K filed October 23 describes a proposed acquisition that would move the small insurance brokerage, which trades under EZRA, into on-demand GPU computing for AI workloads. The target operates a marketplace where third-party GPU owners earn revenue by offering hardware through its platform. Reliance says it intends to acquire additional GPUs and expand marketplace participation after close. The company has previously been focused on insurance distribution. If the acquisition proceeds, it would be a full business pivot into AI infrastructure.
A SPAC Uses AI Power Demand as Its Carbon Capture Thesis
GigCapital9 Corp.'s 8-K filed October 26 announces a proposed business combination with a post-combustion carbon capture technology company. The commercial rationale is framed around AI infrastructure: GigCapital9 describes the proposed combination as supporting "the commercialization of clean, firm power applications required for the growth in power demand driven by artificial intelligence (AI) applications." The AI power demand argument has circulated in infrastructure circles for several years; it is now appearing inside SPAC merger rationales submitted to the SEC.