What public companies just told the SEC about AI, read straight from the filings. Every quote links the original document.
Oura Files S-1 Built on AI Wearable Health Intelligence
Oura Inc. filed its S-1 on September 3, describing the company as "an always-on health intelligence platform" and arguing that chronic conditions rise while care remains episodic, and that AI combined with continuous physiological sensing can close that gap.
Advances in wearable technology and artificial intelligence ("AI") now make that future possible by translating long-term physiological data into predictive insights and actionable guidance so people can understand their bodies, respond earlier, and live better for longer.
Oura is the most prominent consumer health wearable company to file this cycle. Its S-1 frames AI as the structural rationale for the hardware's market value, and the filing's commercial argument rests on that claim entirely.
Eightco Holds $90 Million in OpenAI Through Treasury SPVs
Eightco Holdings (ORBS) disclosed in an 8-K this week that it has invested approximately $90 million in special purpose vehicles with exposure to equity interests in OpenAI, representing 24% of the company's treasury holdings. Worldcoin accounts for another 29%.
Eightco is built around three mega-trends the Company expects to shape the next decade of innovation: artificial intelligence, digital identity, and the creator economy, with positions in each trend through indirect investment in OpenAI (24% of ORBS' treasury holdings), Worldcoin (29%), and Beast Industries (5%).
Eightco is a small-cap holding company. The $90 million OpenAI SPV position makes the stock a vehicle for retail investors seeking indirect private-market AI exposure they cannot access directly.
"Agentic AI" Moves From Marketing Into Formal Risk Factors
Rubrik (RBRK) filed its 10-Q for Q2 2027 on September 1 and named agentic AI explicitly in its risk factors. The company writes that it uses and increasingly relies on both generative and agentic AI tools, including for code generation, and is investing to expand those capabilities through a product it calls Rubrik Agent Cloud.
Our use of generative, agentic and other AI tools may pose risks to business and operations, including our proprietary software and systems, and may subject us to legal liability. We use and increasingly rely upon AI tools in our business, including generative AI and agentic AI, including to generate code and other materials incorporated into our products, proprietary software, and systems...
C3.ai (AI) filed an 8-K on September 2 now naming its core platform the "C3 Agentic AI Platform." ECARX Holdings (ECX) disclosed a 6-K describing WorkBuddy as an agentic AI workbench running on in-vehicle Flyme OS, capable of orchestrating multi-step tasks and generating documents, reports, and code. Tuttle Capital's post-effective amendment registered a fund to be named the Tuttle Capital Agentic AI Income Blast ETF, with an October 2, 2026 effective date. The term now appears in product names, 10-Q risk factors, and ETF registration statements filed within the same week.
Linkhome, a Real Estate Firm, Pivots to 144 NVIDIA GB300 GPUs
Linkhome Holdings (LHAI) filed an 8-K on September 2 announcing the formation of Linkhome Technologies Inc., a wholly owned subsidiary focused on AI computing infrastructure and high-performance computing. The company says it has begun a preliminary evaluation of a European AI computing project that could involve up to 144 NVIDIA GB300 GPUs. Linkhome describes itself as an "artificial intelligence-driven technology company" and began as a real estate technology platform. The filing characterizes the GPU project as a preliminary evaluation with no agreements in place.
Two Filings Show How Far AI Risk Language Has Traveled
Regis Corp (RGS), the hair salon franchise operator, listed its use of artificial intelligence as a forward-looking risk factor in a September 1 8-K. The disclosure carries no elaboration and sits alongside concerns about stylist retention and the company's relationship with Walmart. At this level of specificity, the language reads as a compliance addition with no disclosed operational context.
CYABRA, Inc. filed an amended S-1 on September 1 to list on Nasdaq. The company detects AI-generated disinformation, and its filing describes the threat directly: actors sponsored by foreign governments or competitors who deploy highly sophisticated generative AI tools to generate and spread fake content through bot networks and fake social media accounts. CYABRA is going public as a commercial response to the exact GenAI-enabled capabilities it names in its own risk disclosure.