What public companies just told the SEC about AI, read straight from the filings. Every quote links the original document.
Safe Pro Wins State Department AI Demining Subcontract
Safe Pro Group Inc. (Nasdaq: SPAI) filed an 8-K on August 18 announcing a subcontract under a U.S. Department of State program. The work involves AI-powered demining software for Ukraine, covering threat detection and mapping of landmine contamination.
Safe Pro Awarded Subcontract for Artificial Intelligence Powered Demining Software under U.S. Department of State Program New Award for AI-Powered Threat Detection and Mapping to Support Ukraine's Generational Landmine Crisis
The filing discloses no contract dollar amount. The State Department channel puts this in a different category from a private defense sale, and the Ukraine use case gives it operational urgency that routine defense tech announcements lack.
WhiteFiber Commits $60 Million Cash to AI Data Center Properties
WhiteFiber, Inc. (Nasdaq: WYFI) disclosed a $60.0 million all-cash property acquisition in an 8-K filed August 17, with closing expected in Q4 2026.
The cash purchase price for the properties is $60.0 million. The transaction is expected to close in the fourth quarter of 2026, subject to customary closing conditions.
WhiteFiber describes itself as a provider of high-performance computing data centers with a vertically integrated model built for generative AI workloads. The filing names no seller and identifies no property locations.
"Agentic AI" Appears Across Three Filings in Separate Industries
B-Scada, Inc. (OTC: SCDA) filed a Form 10-12G to register its shares with the SEC after abandoning its prior business entirely and repositioning as an agentic AI company for small and mid-sized corporate clients. The registration statement defines the term on its own pages:
...we have repurposed our business operations to focus on the development and commercialization of agentic artificial intelligence ("AI") solutions tailored primarily for small- and medium-sized business ("SMB") corporate clients. Agentic AI refers to autonomous AI systems capable of performing tasks, making decisions, and interacting with environments or data sources with minimal human intervention, often through multi-step reasoning and tool integration. Our restructured business model positions the Company as an "AI lab" where we design, prototype and deploy bespoke or proprietary...
In a 425 merger announcement, SPAC acquirer Iron Horse Acquisition II Corp. disclosed a target called ELECTRA AI, which describes its product as an "AI Brain for Batteries" platform combining agentic AI, physical AI, and physics-informed battery modeling across grid, energy storage, and other battery-powered sectors.
Weave Communications, Inc. (NYSE: WEAV) disclosed in an 8-K that Francisco Partners is acquiring the company. Weave's filing describes its product as a vertical SaaS platform using agentic AI workflows to handle scheduling, insurance verification, and payment collection for healthcare practices.
SecureTech Reverses Its Spin-Off Plan for AI Unit
SecureTech Innovations, Inc. (OTCQB: SCTH) announced in an 8-K that it will retain its AI UltraProd subsidiary as a permanent, wholly owned business rather than pursue the previously planned independent listing. The stated rationale is accelerating U.S. market expansion under unified leadership. The filing discloses no financial metrics for AI UltraProd.
Boston Properties Lists AI Risk in a Standard Debt Offering
Boston Properties Limited Partnership filed a 424B5 prospectus supplement on August 17 for a debt offering. Its risk factor section now pairs AI risk directly with cybersecurity:
...risks associated with our use of artificial intelligence and cyber security breaches, incidents, and compromises, as well as other significant disruptions of our information technology (IT) networks and related systems, which support our operations and our buildings...
Boston Properties is one of the largest office REITs in the United States. The filing involves no AI product or service. That boilerplate AI risk language now appears in a major real estate company's debt prospectus alongside climate, litigation, and cybersecurity disclosures marks how standard the disclosure has become outside the tech sector.