Why Broad AI Health Platforms Implode: Lessons from Olive, Babylon, Forward

The healthcare AI landscape, once brimming with audacious promises of ubiquitous transformation, is now littered with the remnants of heavily funded, broad-stroke platforms. For investors and health plan executives, the recent implosions of companies like Olive AI, Babylon Health, and Forward Health serve as stark reminders of the perils inherent in pursuing breadth without depth in a sector demanding precision and demonstrable outcomes. These cautionary tales underscore a critical lesson: generalized AI solutions, lacking disease-specific focus and robust clinical evidence, often succumb to unsustainable unit economics and an inability to deliver tangible value.

The High-Stakes Implosion of Horizontal AI Health Platforms

The narrative of horizontal AI health platforms often begins with ambitious claims of disrupting the entire healthcare continuum, only to end in significant value destruction. Olive AI, once valued at $4 billion, ultimately ceased operations, effectively seeing its valuation plummet to $0. Similarly, Babylon Health, which commanded a valuation of $4.2 billion, also met a similar fate, collapsing to $0. Forward Health, another player in the broad-spectrum primary care space, which reached a $1 billion valuation, ultimately shut down in November 2024. These are not isolated incidents but rather a pattern signaling fundamental flaws in their strategic approach.

• Olive AI: Positioned as an AI solution for healthcare administration, Olive aimed to automate a wide array of tasks, from prior authorizations to revenue cycle management. While the vision was expansive, the execution struggled with the inherent complexities and fragmentation of healthcare workflows. The lack of deep specialization meant their AI often skimmed the surface, failing to deliver the transformative efficiency gains promised to healthcare systems.
• Babylon Health: This UK-based company sought to deliver “digital-first” primary care, leveraging AI for symptom checking and virtual consultations. Despite raising substantial capital, Babylon faced persistent questions regarding the clinical efficacy of its AI tools and the sustainability of its business model. Its rapid expansion into multiple geographies and services, rather than focusing on specific disease areas, diluted its impact and strained its resources.
• Forward Health: Offering a membership-based, tech-enabled primary care model, Forward aimed to provide proactive health management through AI and biometric data. While innovative, its high-touch, high-cost model struggled to scale profitably and demonstrate clear, population-level health improvements that would appeal to health plans or broader investor bases beyond early adopters.

Beyond these prominent examples, other platforms have also faced significant challenges. Pear Therapeutics, a pioneer in prescription digital therapeutics (PDTs), which once boasted a valuation of $1.6 billion, saw its value drastically reduced to $27 million before its eventual bankruptcy in April 2023. This outcome, highlighted by sources like CB Insights, illustrates the difficulty even for regulated digital health products to achieve widespread adoption and reimbursement without a highly specialized and clinically validated approach. Companies like Noom, initially celebrated for its behavioral science-backed weight loss program, and Teladoc Health, a telehealth giant, have also experienced significant market corrections and struggles to maintain growth, facing scrutiny over their broad offerings and the depth of their impact across diverse health conditions.

The Diagnosis: Breadth Without Depth, a Fatal Flaw

The common thread weaving through these failures is a fundamental miscalculation: the belief that a generalized AI platform can effectively address the multifaceted and highly specialized challenges of healthcare. This “breadth without depth” approach manifests in several critical shortcomings:

• Missing Clinical Evidence and Outcomes Data: Healthcare demands rigorous validation. Horizontal platforms often dilute their focus across too many conditions or administrative tasks, making it difficult to generate the robust, disease-specific clinical evidence required to prove efficacy and secure reimbursement. Without clear, outcomes-data-supported proof points, adoption by risk-averse health systems and payers remains elusive.
• Unsustainable Unit Economics: Scaling a generalized platform across diverse healthcare needs often leads to high operational costs without commensurate revenue generation. The investment required to build, deploy, and maintain AI for a vast array of use cases, each with its own unique data requirements and regulatory hurdles, far outweighs the value delivered when that AI lacks deep, specialized impact.
• Lack of Domain Expertise Integration: Effective AI in healthcare is not just about algorithms; it’s about embedding those algorithms within deep clinical and operational workflows. Horizontal platforms frequently struggle to integrate the nuanced domain expertise required for specific diseases or operational challenges, leading to solutions that are technically impressive but clinically irrelevant or impractical.
• Regulatory and Reimbursement Labyrinth: Navigating the regulatory landscape (e.g., FDA 510(k) clearance or De Novo classification) and securing reimbursement (e.g., CPT codes, NTAP) is arduous, even for highly specialized solutions. For platforms attempting to cover a multitude of conditions or services, this becomes an exponential challenge, often resulting in “evidence-light” offerings that fail to meet the stringent requirements of payers and providers.

As Rock Health data consistently shows, while digital health funding has seen an AI-powered rebound, reaching $7.4 billion in the first half of 2026, with capital concentrating in a smaller number of firms and megadeals, the ability of many companies to translate that funding into sustainable, impactful businesses has been limited, particularly for those lacking clear specialization.

Expert Commentary on the Horizontal Pitfalls

The challenges faced by these broad AI platforms have not gone unnoticed by industry stalwarts. Dr. Eric Topol, a renowned cardiologist and leading voice in digital medicine, has consistently emphasized the need for AI in healthcare to be clinically validated and integrated thoughtfully, often cautioning against the hype surrounding generalized solutions that lack robust evidence. His commentary frequently points to the importance of deep, disease-specific understanding for AI to truly augment human intelligence in medicine. Eric Topol’s commentary on AI in medicine Casey Ross, an investigative reporter at STAT News, has extensively covered the digital health sector, often highlighting the financial struggles and operational missteps of companies that promised sweeping transformations but failed to deliver. Ross’s reporting has frequently delved into the unsustainable financial models and the lack of demonstrable outcomes that plagued many of these heavily funded, yet ultimately unsuccessful, horizontal platforms. His analyses, often grounded in financial and operational realities, provide a critical perspective on why these ventures falter. Casey Ross’s investigative reporting on digital health The consensus among these authoritative voices is clear: the pursuit of broad impact without deep, evidence-based specialization is a recipe for failure in healthcare AI.

Lessons for Strategic Investment and Health Plan Integration

The spectacular failures of Olive AI, Babylon Health, and Forward Health are not merely isolated incidents; they are critical data points for investors and health plan executives charting the future of healthcare innovation. The overwhelming evidence points to a strategic imperative: prioritize vertical AI healthcare companies and disease-specific AI health platforms. For investors, this means a rigorous due diligence process that moves beyond market hype and focuses on:

• Clinical Validation: Insist on robust, outcomes-data-supported clinical evidence for specific disease states. Look for companies that have invested heavily in generating real-world evidence (RWE) and demonstrating measurable improvements in patient outcomes or operational efficiencies within a narrow, defined scope.
• Regulatory Clarity and Reimbursement Pathways: Companies with clear FDA clearances (e.g., 510(k), De Novo) and established or emerging CPT codes for their specialized interventions demonstrate a far more viable path to market. A well-defined reimbursement strategy, often tied to specific diagnostic or therapeutic procedures, is paramount.
• Deep Domain Expertise: Invest in teams that possess profound clinical and technical expertise within their chosen vertical. An AI health vertical specialization in areas like behavioral health, oncology, or diabetes, built by experts in those fields, is far more likely to yield impactful and sustainable solutions.

For health plan executives, the lesson is equally profound. Partnering with disease-specific AI platforms offers a more direct and measurable route to improving member health and managing costs. Instead of seeking a single, all-encompassing solution, health plans should:

• Target Specific Disease Burdens: Identify high-cost, high-prevalence conditions within their member populations (e.g., diabetes management, behavioral health support, early oncology detection). Seek out AI solutions purpose-built to address these specific challenges, such as a dedicated behavioral health specialization AI tool.
• Demand Outcomes-Based Contracting: Engage with vertical AI companies that are confident enough in their specialized outcomes to enter into value-based agreements. This aligns incentives and ensures that the technology delivers tangible results for members and the plan.
• Integrate Thoughtfully: Specialized AI tools often integrate more seamlessly into existing clinical workflows because they are designed with those specific workflows in mind. This reduces implementation friction and increases adoption rates among providers.

The trajectory of these horizontal platforms serves as a stark reminder that in healthcare AI, depth of impact within a focused vertical consistently outperforms the allure of broad, generalized solutions. The path to sustainable value creation lies in precise, evidence-backed specialization.