BSUM launches AI-native due diligence room Haihtuva

BSUM has launched Haihtuva, an AI-native due-diligence data room for investments and M&A, with first paid pilots opening today. The platform is designed to keep files in the seller’s own repository while giving deal teams AI-driven review, risk flagging, and Q&A workflows.

Why it matters: - Haihtuva targets a core pain point in dealmaking: how to use AI on sensitive documents without moving them into a separate system. - The product could appeal to buyers, investors, sellers, and target companies that want faster diligence with tighter confidentiality controls. - BSUM is positioning the platform as a new model for virtual data rooms, not just another add-on AI tool.

What happened: - BSUM (Biotech Start-Up Management Oy Ltd) announced Haihtuva on Oct. 9, 2026, in Helsinki. - Haihtuva is an AI-native due-diligence room built for investments and M&A dealmaking. - The first paid pilots for Haihtuva opened the same day. - The product is now available at Haihtuva.

The details: - Haihtuva is designed from the ground up as an AI-native virtual data room, rather than a legacy VDR with AI added later. - For sellers, Haihtuva runs readiness checks and gap analysis against the seller’s own data room. - For buyers, Haihtuva automates document-level risk flagging and question-and-answer workflows. - Documents are never uploaded to Haihtuva. - The platform reads documents on demand, read-only, directly from the seller’s or target company’s own repository. - The buyer or investor chooses which AI model and training data run against those documents. - Processing happens in volatile memory inside confidential, attested computing environments. - The memory is cleared after the work is done. - BSUM says the processing approach is covered by patent applications. - Haihtuva includes standard VDR functions, including granular read-permission settings and usage reporting. - Teams can also use traditional human review instead of AI workflows.

Between the lines: - BSUM is betting that confidentiality, not just automation, will be the main selling point for AI in diligence. - The “volatile” design and buyer-controlled AI model choice are meant to answer concerns about data exposure and model governance. - The company is also drawing a sharp contrast with older VDRs that rely on document indexing systems built years ago. - Juhani Lahdenperä, BSUM co-founder, said the goal was to let transactions happen without client files ever being uploaded outside the client’s organization. - Lahdenperä also said the platform was built to put AI into the room from the start rather than bolt it onto an older document index. - BSUM says the team behind Haihtuva brings more than 25 years of hands-on life science due-diligence experience.

What’s next: - BSUM is opening Haihtuva to its first paid pilots. - The company appears to be using pilot customers to validate the product’s AI-native workflow and confidentiality model before wider rollout. - If adoption grows, Haihtuva could compete for deals where regulated documents and privacy concerns slow down AI use.

The bottom line: - Haihtuva is BSUM’s attempt to make AI due diligence work without surrendering control of sensitive files.