What public companies just told the SEC about AI, read straight from the filings. Every quote links the original document.
Vulcan Closes $39 Million to Scale AI Power Infrastructure
Vulcan Infrastructure and Power (Nasdaq: VIP) closed a $39.4 million strategic investment from affiliates of Machine Investment Group, Atlas Holdings, and Conversant Capital. The 8-K describes more than 100 MW of "immediate and near-term AI/HPC opportunities" and a 654 MW development pipeline across owned sites. Vulcan describes itself as a power infrastructure platform built around acquiring and operating energized sites that host AI and high-performance computing data centers.
PowerCompute (Nasdaq: PWCM) reported its August operations separately, noting 323 Bitcoin in treasury valued at approximately $25.2 million and $132,000 in energy sales from curtailment. Its filing describes the company as "a Bitcoin treasury and mining company expanding into high-performance computing and artificial intelligence infrastructure." The AI pivot is still early, but the capital position and the energy infrastructure context appear in the same sentence for a reason.
Three Companies File "Agentic AI" as Their Business Identity
The S-1 registration filed this week by OLB Group (Nasdaq: OLB), a payments and commerce company, frames agentic AI as a structural commitment rather than a product feature:
We have been re-engineering our operations around artificial intelligence ("AI"), including agentic AI — autonomous software agents that are designed to plan, decide and execute tasks with limited human intervention. Our objective is to operate as an AI-native financial technology company.
That language comes before financial statements, not after them. The S-1 is a public declaration of business identity, and OLB's is built around this framing.
RUM Group, parent of the Rumble video platform, filed an 8-K describing its infrastructure buildout as "building the rails of the agentic-first enterprise: the AI compute, cloud infrastructure, and trust layer for the agentic AI future." That language sits alongside descriptions of creator video hosting in the same document.
Amber International Holding listed "our business transformation into an agentic AI company" as a forward-looking priority in a 6-K, while the same filing separately flagged the company's intent to "successfully remediate the material weaknesses in our internal controls." The combination is worth noting.
NICE Discloses Acquisition of a Named Agentic AI Platform
NICE Ltd. disclosed the acquisition of NiCE Cognigy in a 6-K filing covering its 2025 sustainability report. The filing identifies the acquisition by name and purpose:
The acquisition of NiCE Cognigy, a market-leading agentic AI platform, marked a significant milestone in strengthening NiCE's position as a leader in enterprise AI.
NICE also reported reinvesting 14% of revenue in research and development in 2025, led by a global team of more than 3,500 R&D professionals.
Build-A-Bear and Macy's Both Added AI Risk Language in Quarterly Filings
Macy's 10-Q for the quarter ended August 1 lists "business, legal and ethical challenges related to our use of artificial intelligence in our business operations" among its risk factors. Build-A-Bear Workshop filed a more detailed version in its own 10-Q:
...our use of artificial intelligence technologies presents operational, reputational, data security and legal risks that could adversely affect our business and financial performance, and any failure to effectively leverage artificial technologies in our business could negatively impact our customer engagement and competitive position
The Build-A-Bear language captures two risks in the same clause: getting AI wrong, and failing to use it at all. That pairing is now appearing regularly in quarterly filings from consumer companies that have no dedicated AI product line. Braze and Navan filed substantively similar language in their 10-Qs this week, adding to the pattern without adding anything specific to the facts about their actual AI use.