AI-powered blood tests hit FDA fast track, fuel funding surge

A clear regulatory marker arrived before the late-September crescendo: The Manila Times reported the headline, “Gene Solutions’ SPOT-MAS 10 Wins FDA Breakthrough Tag for Multi-Cancer Blood Test Covering 10 Cancers.” That matters because FDA Breakthrough designation is not marketing language but a formal signal that multi-cancer blood tests had moved into a more serious regulatory lane, reinforcing the broader late-2026 picture in which AI-powered liquid biopsy platforms were attracting major capital and institutional attention at the same time regulators were beginning to confer high-profile validation.

Just as important, not every high-visibility announcement counted as proof of that shift. The provided material is a late-October 2026 promotional campaign for Caris Detect, not evidence of late-September 2026 funding or FDA breakthrough designations: a PR Newswire release dated “Oct. 19, 2026” announcing “Caris Life Sciences Launches New National Commercial… featuring… Jordan Spieth” to promote “Caris Detect™, Caris' multi-cancer early detection (MCED) test.” That distinction helps isolate the real market signal—regulatory milestones and financing—from later brand advertising around the category.

The technical shift is from hunting one biomarker at a time to letting AI read many weak clues together as a composite cancer signal. As NEXT with John Koetsier described it, early disease is hard because “the earlier that you try to detect, the less signal,” so the answer is to “fuse DNA,... gene expression,... methylation” and let AI decipher across all those sensors at once, more like triangulating with multiple satellites than trusting a single reading from a single molecular channel.

That fusion logic already shows up in adjacent oncology workflows: Where Tech Meets Bio reported that Microsoft Research, Providence Health, and the University of Washington built GigaTIME, which “routinely converts pathology slides into virtual immune-protein maps,” surfacing over 1,200 significant associations across 14,256 patients, while the same outlet noted Linköping University’s “32-sensor electronic nose” could distinguish ovarian cancer in blood plasma with 97% accuracy in ten minutes—no biomarker panel, just pattern recognition across volatile compounds. Together, those examples show why integrating cfDNA, RNA, methylation, and protein from one blood draw can overcome fragmented testing and expose faint early-stage signals.

The clearest evidence of measurable gains comes from disease-specific validation data rather than broad promises. In Breast Cancer: Targets and Therapy, the study “Development and Validation of a Machine-Learning Deep Plasma Proteome Classifier for Early Stage Breast Cancer Detection” reported a sensitivity of 92.6% and specificity of 92.3%, while the validation arm of 397 women showed results consistent across all breast cancer stages (0-IV) and across various molecular and pathological subtypes, including Invasive Lobular Carcinoma (ILC).

The evidence also extends beyond headline accuracy into harder clinical and operational tests of usefulness. For women with dense breasts, Certitude detected 93% of breast cancers in women with dense breast tissue across a simulated population, and the 99% completion rate for Shield in two rural South Carolina clinics highlights how a simple blood draw can address screening gaps in medically underserved communities. The COSMOS MEL01 data for Guardant Reveal in stage II III melanoma shows how Guardant Health is extending blood based testing from screening into recurrence monitoring, and in May 2026 the FDA approved an immunotherapy after surgery for patients whose blood still showed molecular residual disease, as determined by an authorized test.

By late September 2026, the market signal around AI diagnostics, including liquid biopsy and MCED, looked less like speculative biotech enthusiasm and more like a capital-intensive precision-oncology land grab. Yahoo Finance described AI-driven liquid biopsy as a “clinically deployable, commercially scalable” tool, while broader market analysis showed why investors were willing to pay up: private U.S. investment in AI reached $109.1 billion in 2024, U.S. digital health funding totaled $10.1 billion, and 63% of rounds still sat at Seed through Series B, indicating a deep pipeline behind the category’s next wave.

That capital was reinforced by regulatory and commercial signals that lowered perceived execution risk. The FDA approved 223 AI-enabled medical devices in 2023, and Yahoo Finance noted that breakthrough-style validation had become central to the investment case; in adjacent oncology AI, Paige’s FDA Breakthrough Device Designations for Paige Lymph Node and Paige PanCancer Detect directly reduced regulatory risk, while Proscia raised $50 million in March 2025 and laboratories were “no longer evaluating AI as a future capability — they are deploying it now,” with PathAI extending its Labcorp partnership in 2026.