Blog

Perplexity

The most notable recent news in this segment is that on August 5, 2026, the FDA approved the first drug that addresses the full spectrum of symptoms for type 1 narcolepsy. This is an important signal for the AI-biomedical market, not only as a clinical event but also as confirmation that new therapeutic platforms are reaching practical application faster.

The same weekly report also noted the approval of a new viral immunotherapy for melanoma, but for a separate news item on AI in medicine, the stronger angle is precisely the acceleration of bringing innovative drugs into clinical practice. Against the backdrop of August deals and partnerships in biopharma, this shows that AI is increasingly influencing not abstract research, but concrete regulatory and therapeutic outcomes.

Sources:

• Healthcare AI & Digital Health Weekly Report · News & Updates · August 10, 2026.

Gemini

New AI Framework Accelerates Search for Therapeutic Antibodies

Researchers from Boston University have developed a specialized AI framework that significantly enhances the efficiency of searching for therapeutic antibodies. Unlike more general models, the new approach focuses on key antibody regions responsible for target binding. This allows for more accurate prediction of binding strength and reduces the computational resources needed to identify the most promising candidates even before laboratory testing begins. The study is published in the journal Communications AI & Computing.

AI Reduces Drug Development Timelines to Months

According to a Forbes report, generative AI is revolutionizing the drug discovery process, reducing timelines from years to months. This acceleration requires pharmaceutical companies and clinical operators to adapt their strategies. The integration of AI also presents new challenges for other industries, such as robotics manufacturing and energy system management.

New Method Enhances AI Reliability in Medical Diagnostics

Researchers from Johns Hopkins University have introduced a new method that significantly improves the reliability and accuracy of AI applications in medicine. The MIGHT (Multidimensional Informed Generalized Hypothesis Testing) method was developed to ensure the high level of confidence required for clinical AI applications. As an example, researchers applied it to the early detection of cancer from blood samples (liquid biopsy), as well as to the analysis of DNA fragmentation, which can indicate not only cancer but also autoimmune and vascular diseases.

Sources:

• Communications AI & Computing
• Forbes
• Johns Hopkins Medicine

ChatGPT

In July 2026, the application of artificial intelligence (AI) in medical practice was expanded in Russia. Doctors can now use AI not only for mammography, X-ray, fluorography, and computed tomography of the chest organs, but also for electrocardiography (ECG) and colonoscopy. This leads to increased diagnostic accuracy and faster diagnosis for patients. (gazeta.ru)

In May 2026, at the St. Petersburg International Economic Forum, the prospects for the development of sovereign AI in medicine were discussed. It was noted that by 2030, the volume of the Russian medical AI market could reach 78 billion rubles, which is more than six times the current figure. Growth is supported by national projects, integration with state infrastructure, and import substitution policies. (expert.ru)

In June 2026, the Ministry of Health of the Russian Federation approved the export of medical standards and the regulation of medical AI systems. This decision opens up new opportunities for domestic developers in the field of medical technologies and contributes to their expansion into international markets. (axioma-ai.ru)

Grok

August 5–6, 2026 saw researchers from Stanford (Brian Hie and colleagues) design complete functional genomes of bacteriophages—viruses that infect bacteria—for the first time using the generative models Evo1 and Evo2. Out of hundreds of generated variants, 16 viable phages with significant evolutionary novelty were produced in the lab; they successfully infected E. coli, including antibiotic-resistant strains.

This is a milestone in synthetic biology: AI has moved from designing individual genes or molecules to creating entire replicating genomes. The technology paves the way for personalized phage therapy against resistant infections, although it simultaneously raises biosafety and control concerns.

The work is published in Science and is already being discussed as a proof of concept for "genomic design" using large language models of biological sequences.

Claude

Found important news. I'll clarify further on recent events from July-August. Great, I've found fresh information. I'll provide an answer in Russian, focusing on recent developments.

---

In the first quarter of 2026, digital health startups raised $4 billion in venture capital, a