Anthropic and OpenAI Step Up Biotech M&A Strategy as AI Deepens Its Footprint in Drug Discovery

The race to integrate artificial intelligence deeply into pharmaceutical research and development reached a significant milestone as leading foundational model providers accelerate their investments in biotech infrastructure. Anthropic announced on October 27 that it acquired Mithril Medicine, an early-stage startup focused on developing experimental treatments for two rare forms of muscular dystrophy. The acquisition highlights a broader industry trend where artificial intelligence developers are moving past software partnerships to acquire specialized biotechnology firms, testing their models on real-world drug programs from inception to preclinical trials.

Mithril Medicine represents a unique case study in lean, AI-driven biotechnology operations. Founded by Cody Tranbarger, who served as the startup’s sole full-time employee, Mithril relied heavily on Anthropic’s Claude model to coordinate a complex operational matrix. Across 18 months, the company successfully took a lead therapeutic candidate from an initial idea to promising preclinical results in mice. Tranbarger utilized Claude to manage logistics, organize legal contracts, and oversee scientific experiments distributed across 11 external laboratory partners, effectively delegating core administrative and operational components of the business to the language model.

By bringing Mithril in-house, Anthropic gains more than just potential intellectual property and therapeutic candidates. Running authentic drug development programs provides the company with empirical testing grounds to evaluate where its Claude model excels and where it encounters limitations in complex scientific workflows. Furthermore, these internal programs generate concrete case studies that Anthropic can share with prospective biotech partners to demonstrate practical utility in high-stakes regulatory and scientific environments. Following the acquisition, Tranbarger has officially joined Anthropic and outlined plans to initiate several additional drug programs.

This transaction follows Anthropic’s earlier maneuvers in the life sciences sector. In April, PYMNTS reported that Anthropic acquired Coefficient Bio, a specialized startup consisting of fewer than 10 employees, for approximately $400 million. Coefficient Bio had built proprietary software tools designed to streamline the planning of drug research and manage intricate clinical regulatory strategies.

Despite the rapid progress demonstrated by Mithril’s lead antibody, Anthropic emphasized that the therapeutic work remains strictly in the preclinical stage, having been tested exclusively in murine models. As these pipeline programs advance, the company intends to extend Claude’s application capabilities into formal clinical trials. Current internal projections suggest that a viable drug candidate stemming from these initiatives could be ready to enter human clinical trials as early as late 2028.

OpenAI Buys Orchestra Bio to Track Drug Workflows

While Anthropic focuses heavily on hands-on therapeutic development and preclinical execution, rival model provider OpenAI is targeting a different operational bottleneck within the pharmaceutical pipeline: the orchestration and tracking of complex research workflows. On October 20, OpenAI announced its agreement to acquire Orchestra Bio, a digital platform designed to allow scientists to delegate study monitoring, milestone oversight, and timeline tracking to autonomous artificial intelligence agents.

The administrative burden Orchestra Bio targets is substantial. Moving a therapeutic candidate from initial target discovery through rigorous preclinical testing to final regulatory approval is notoriously arduous, typically spanning a timeline of 10 to 15 years, as noted in previous industry reports. Orchestra Bio’s specialized agents continuously monitor studies and operational timelines across an entire drug program.

For instance, if a critical manufacturing batch fails during production and threatens to delay the scheduled launch of a clinical trial, Orchestra’s agents analyze the disruption. The system automatically identifies which downstream activities and critical dates are affected, retrieves the relevant documentation, and lays out strategic operational options for human researchers. Program leaders and scientists retain ultimate authority to evaluate the data and decide on the appropriate course of action.

Orchestra Bio will integrate into OpenAI’s growing ecosystem, joining Rosalind, a dedicated suite of AI tools commercialized specifically for drug researchers. Major pharmaceutical and life sciences leaders, including Amgen, Moderna, and Thermo Fisher Scientific, are already collaborating with OpenAI to utilize its specialized GPT-Rosalind model. Prior to its acquisition by OpenAI, Orchestra Bio operated as a four-year-old startup that had successfully raised approximately $20 million from prominent venture capital backers, including Andreessen Horowitz, Atria, and 8VC.

Enveda Cuts 26 Hours of Batch-Record Review to Three

While multi-year clinical trials and therapeutic discovery capture widespread attention, immediate operational cost savings are materializing in regulatory paperwork and documentation review. Enveda, a biotechnology company that leverages artificial intelligence to mine nature’s chemistry for novel pharmaceutical compounds, partnered with Anthropic to tackle the exhaustive manufacturing records that drug developers must meticulously file prior to launching clinical trials. Historically, Enveda’s trained chemists had been spending up to 75% of their working hours wrestling with dense administrative documents.

To alleviate this burden, Enveda developed a specialized tool powered by Claude designed to review manufacturing batch records. In benchmark testing, the Claude-backed tool successfully caught all 11 compliance issues that human reviewers had previously identified across a massive 1,061-page document set. More importantly, the AI tool proactively flagged seven additional issues that human reviewers had initially missed. Crucially, the deployment reduced the total document review time from nearly 26 reviewer-hours down to just under three hours, representing an 89% reduction in labor time for this specific operational workflow.

Healthcare and biotechnology buyers are actively budgeting for enterprise artificial intelligence solutions to achieve similar efficiencies. According to findings from the PYMNTS Intelligence report titled “Enterprise AI Gains Ground, but Scale Remains Uneven,” which surveyed 60 senior technology executives at United States enterprises generating at least $1 billion in annual revenue, 84% of surveyed healthcare firms stated they expected to increase their overall artificial intelligence spending over the ensuing 12 months, with zero respondents anticipating a reduction in tech budgets.

This surging enterprise demand aligns closely with broader corporate strategies. In the S-1 registration statement Anthropic filed on October 26, the company explicitly listed “healthcare and life sciences” as the primary new vertical market it intends to aggressively target. This push occurs within an expansive global sector where total pharmaceutical spending is projected to surpass $2 trillion. According to the regulatory filing, nine of the ten largest pharmaceutical corporations in the world are already active customers, though the document does not publicly disclose their names. Furthermore, company executives have previously indicated to potential investors that expanding its biological and healthcare applications will enhance public perception and trust in artificial intelligence technologies ahead of an initial public offering.