FDA

The FDA has a clear regulatory framework for AI-assisted medical decision-making tools, classifying them as Software as a Medical Device (SaMD).

2023年FDA发布了AI/ML药物开发讨论文件，但仅是初步探索

The FDA's Adverse Event Reporting System (FAERS) receives approximately 500,000 to 600,000 adverse event reports per quarter.

美国FDA已于2023年发布专项指南，认可合成医疗数据在特定条件下的监管有效性

School food colorant bans promoted at the state level such as in California, and the FDA's revocation of authorization for Red No. 3 (erythrosine), form the policy backdrop for companies' formula adjustments.

The United States operates two systems for food colorants, "certified colorants" and "colorants exempt from certification"; synthetic colors such as Brilliant Blue, Tartrazine and Sunset Yellow belong to the former and require FDA batch certification, while natural-source spirulina extract falls under the latter.

Hard-tech companies typically involve prototyping, supply chains, long R&D cycles, and high capital expenditure, while life sciences must go through preclinical research, clinical trials, and regulatory approval (such as FDA), with cycles often measured in years.

Regulators such as the FDA and CE have begun requiring submission of subgroup performance analyses to ensure that AI tools are safe and effective across different populations.

In the United States, the FDA reviews animal genome editing as an "animal drug" under the Federal Food, Drug, and Cosmetic Act, a slow and costly process; the USDA has a separate evaluation pathway for some agricultural animals.

The U.S. market has relatively strict regulation of wearable health devices, and products must pass FCC certification and FDA-related compliance requirements.

Regulators (such as the U.S. FDA and China's NMPA) typically require prospective data, or at least rigorously designed multicenter retrospective validation, for high-risk diagnostic AI.

BPC-157 has not yet received formal approval from mainstream regulatory agencies (such as the FDA) for clinical treatment.

GxP is the collective term for the Good Practice series of regulations, covering GMP, GCP, GLP, and others, enforced by national drug regulatory agencies (such as the U.S. FDA and China's NMPA).

In the United States, the FDA regulates products claiming to treat, diagnose, or mitigate diseases as medical devices; if a product is used only for general wellness and self-management, it is generally not classified as a medical device, but marketing language must avoid disease-related claims.

Regulators such as the FDA and EMA increasingly require mechanistic justification for AI-driven diagnostics.

The US FDA classifies some clinical decision support software (CDS) as medical devices, but software may be exempt if it is used only to display information and medical professionals can independently review the basis for it.

Medical-grade exoskeleton devices used for rehabilitation training or mobility assistance are classified as medical devices in most markets and require approval by drug regulatory authorities; in China they are Class II or Class III medical devices under NMPA, in the United States they follow the FDA's 510(k) or De Novo pathway, and the clinical validation cycle typically takes years.

In the United States, consumer health apps that claim to diagnose, treat or prevent disease may fall within the FDA's definition of medical devices; in China, services providing psychological assessment and intervention must comply with regulations on mental health and internet-based medical care.

The Prima system is submitting an application to the US FDA and awaiting marketing authorization, but the specific timeline is not yet clear.

FDA approval of high-risk medical devices usually requires animal experiment data to demonstrate safety, biocompatibility, and functional effectiveness.