On January 6, 2026, the U.S. Food and Drug Administration (“FDA”) published revised final guidance documents on Clinical Decision Support Software (“Revised CDS Guidance”) and its General Wellness Policy for Low Risk Devices (“Revised General Wellness Guidance”). FDA Commissioner Martin Makary announced the revised guidance documents in a video posted on X and in public remarks at the Consumer Electronics Show (“CES”), stating that FDA needs to “adapt with the times” and touting the changes as promoting innovation and capital investment in artificial intelligence (“AI”) and digital wearable technologies. Commissioner Makary also announced that FDA is in the process of developing a “new regulatory framework for AI” that is “smarter and more forward-thinking.”
The specific changes FDA has implemented in the Revised CDS Guidance and Revised General Wellness Guidance address concerns about overregulation raised by industry. In general, they expand the scope of digital health products that FDA will not regulate, either due to FDA’s interpretation of the scope of its legal authority or as a matter of FDA policy. This Alert analyzes the key changes to the guidance documents, Commissioner Makary’s broader remarks regarding AI and digital health product regulation, and the implications for product developers and other digital health stakeholders.
General Wellness Guidance
FDA first published guidance addressing its enforcement policy for low-risk general wellness products as a draft in 2015, which it finalized in 2016 and subsequently revised in 2019 following statutory changes enacted in the 21st Century Cures Act. Products falling within the scope of this policy can be marketed without FDA oversight. Prior Ropes & Gray Alerts have discussed the 2015 draft and 2016 final versions of the General Wellness Guidance.
The most recent revision leaves the basic policy framework from earlier versions of the General Wellness Guidance intact, continuing to define general wellness products subject to the policy as those (1) intended for only “general wellness” use and (2) presenting a low risk to the safety of users and other persons. However, the Revised General Wellness Guidance significantly expands the scope of products that FDA will consider as being intended for general wellness uses. In particular, the revised guidance opens the door for non-invasive wearable products that measure physiologic parameters, including blood pressure and blood glucose, to be considered general wellness products when marketed for general wellness purposes. Commissioner Makary asserted in his CES remarks that this change creates “clear lanes” for “consumer grade” products versus “medical grade” products, thereby reducing the amount of “subjectivity by regulators and guesswork by developers.”
The guidance clarifies that FDA “may consider certain products that use non-invasive sensing (e.g. optical sensing) to estimate, infer, or output physiologic parameters (e.g. blood pressure, oxygen saturation, blood glucose, heart rate variability) to be general wellness products when such outputs are intended solely for wellness uses, and meet certain criteria.” Those criteria are that the product:
- is non-invasive and not-implanted;
- does not pose a safety risk if specific regulatory controls are not applied;
- is not intended for the diagnosis, cure, mitigation, prevention, or treatment of a disease or condition;
- is not intended to substitute for an FDA-authorized, cleared, or approved device;
- does not include claims, functionality, or outputs that prompt or guide specific clinical action or medical management; and
- does not include values that mimic those used clinically unless validated (e.g. manufacturer testing, peer-reviewed clinical literature) to reflect those values.
The guidance also notes that products meeting these criteria may display data values, ranges, trends, baselines, or longitudinal summaries, and may contextualize these outputs in relation to sleep, activity, stress, recovery, or similar wellness domains. In addition, these products may inform a user that evaluation by a healthcare professional may be helpful when outputs fall outside ranges appropriate for general wellness use, so long as they do not identify a specific medical condition, characterize the output as abnormal or diagnostic, identify clinical thresholds, or provide ongoing alerts or monitoring intended to manage a disease or condition.
The revised guidance incorporates two new examples of low risk general wellness products to illustrate this expanded scope. They are:
- A wrist-worn wearable product intended to assess activity and recovery that outputs multiple biomarkers, among which are hours slept, sleep quality, pulse rate, and blood pressure, where the claims exclusively relate to general wellness.
- A non-invasive wearable product advertised toward elite athletes and intended for monitoring several parameters, among which are electrolyte balance, lactate, and hemoglobin, where the product is labeled for use in an exercise/fitness context only, displays values from cleared devices or ranges from the wearable’s optical sensor, and is disclaimed for use in diagnosing any condition or disorder.
The Revised General Wellness Guidance represents an important change in FDA’s position on wearable products that measure and display physiologic parameters like blood pressure and blood glucose. Most notably, in July 2025, FDA issued a warning letter to WHOOP, Inc., asserting that a blood pressure feature in the company’s wearable product could not fall within the agency’s general wellness policy because it was “inherently associated” with the diagnosis of hypo- and hypertension. That position is incompatible with FDA’s Revised General Wellness Guidance, which clarifies that whether a non-invasive wearable can be considered a general wellness product depends on how its manufacturer or distributor advertises and promotes it, not whether it provides an output that is “inherently associated” with a disease or condition.
The Revised General Wellness Guidance will support an expansion of the measurements and data provided by wearable products in the future. As Secretary of Health and Human Services, Robert F. Kennedy, Jr. said on X after FDA published the Revised General Wellness Guidance, FDA “is unlocking wearable devices for general wellness that have, for too long, been tied up in red tape . . . [and] can help everyday Americans proactively take charge of their health, helping them steer clear of expensive interventions and chronic disease.” However, the manner in which FDA implements the new policy will be crucial, and developers of products that measure physiological parameters with well-known clinical implications should continue to exercise great care. Developers choosing to bring such products to market as low risk general wellness products should pay close attention to the criteria FDA has set out, particularly with regard to product advertising and promotion. Also, under the Revised General Wellness Guidance, FDA states that products providing “values that mimic those used clinically” should be “validated to reflect those values.” The guidance does not define what the values that fall into this category may be, what that validation should look like, or how FDA would be able to apply this aspect of the guidance. Nonetheless, companies developing general wellness products that provide measurements that may also have clinical uses should ensure that they have documented validation to support those measures, as such validation will not only align with FDA’s expectations but may also help guard against legal risks under federal and state consumer protection laws.
CDS Guidance
Like the General Wellness Guidance, FDA has revised the CDS Guidance multiple times, issuing draft guidance in 2017, revised draft guidance in 2019, and final guidance in 2022. Prior Ropes & Gray Alerts have discussed the 2022 CDS Guidance, as well as FDA’s 2025 responses to citizen petitions challenging the CDS Guidance. Each of the prior iterations of the CDS Guidance represented a substantial overhaul in the agency’s approach. The latest Revised CDS Guidance instead retains the core approach from the prior version, but includes targeted yet important updates. These include a new enforcement discretion policy for CDS tools that provide only a single recommended treatment option and removal of language from the earlier guidance stating that software providing a risk score or risk probability does not qualify as non-device CDS. In his CES remarks, Commissioner Makary described how FDA’s prior stance on these aspects of CDS in the 2022 guidance had “unintended consequences,” including that it had “forced” developers to “build dumber software,” claiming that the revised guidance “stops the perverse incentive to build dumber software, and will bring hundreds of millions of dollars of investment off the sidelines.”
The Revised CDS Guidance, like the prior iterations, provides FDA’s interpretation of the statutory criteria for CDS software to be exempt from FDA regulation as a medical device. Specifically, pursuant to Section 520(o)(1)(E) of the Food, Drug, and Cosmetic Act (“FDCA”), CDS software is excluded from the definition of a device if it meets all four of the following criteria:
- Is not intended to acquire, process, or analyze a medical image or a signal from an in vitro diagnostic device or a pattern or signal from a signal acquisition system (Criterion 1);
- Is intended to display, analyze, or print medical information about a patient or other medical information, like clinical practice guidelines (Criterion 2);
- Is intended to support or provide recommendations to an HCP about prevention, diagnosis, or treatment of a disease or condition (Criterion 3); and
- Is intended to enable HCPs to independently review the basis for the software’s recommendations so HCPs do not primarily rely on the recommendations when making a clinical diagnosis or treatment decision (Criterion 4).
The most substantive changes in the Revised CDS Guidance are to FDA’s interpretation of Criterion 3. In the 2022 version of the guidance, FDA had expressed concern that “automation bias” may be more likely to occur if software provides a user with a single, specific, selected output or solution rather than a list of options, and it interpreted Criterion 3 as encompassing only CDS software that does not provide a specific preventive, diagnostic, or treatment output or directive. Although FDA has not changed its legal interpretation, in the new version of the guidance, FDA has added an enforcement discretion policy to encompass CDS that meets all the criteria for non-device CDS, but that only provides a single recommendation that is clinically appropriate. Examples provided in the guidance include software functions that predict risk of future cardiovascular events based on patient medical information, create a recommended treatment plan for patients diagnosed with cognitive impairment, or recommend a specific antibiotic for an HCP to consider based on patient symptoms and other inputs. Another example is a software function that provides an HCP with a differential diagnosis, which may present either multiple diagnostic considerations or a single clinically appropriate diagnostic recommendation when alternative diagnoses are “highly improbable.” FDA notes that recommendations from the software should be reviewed, revised, and finalized by an HCP.
In addition, the 2022 version of the guidance included a statement that “FDA considers software that provides information that a specific patient ‘may exhibit signs’ of a disease or condition or identifies a risk probability or risk score for a specific disease or condition” as failing Criterion 3, and thus subject to FDA regulation. This narrow interpretation received significant pushback from industry stakeholders as inconsistent with the statutory criteria for exemption and unduly restrictive of innovation in digital health product development. FDA has now removed that language from the guidance and added several examples of CDS software functions that provide risk scores (e.g., for cardiovascular events or post-operative complications) or differential diagnoses, which now would qualify for enforcement discretion under the revised policy.
The Revised CDS Guidance also modifies where and how FDA addresses time criticality and automation bias as relevant factors in determining whether a software function will be regulated, although the intent and effect of these revisions is not entirely clear. FDA revised the body of the guidance document to address these concerns in relation to Criterion 4 rather than Criterion 3, apparently in response to concerns that had been raised that these factors were not relevant to Criterion 3. Yet in the illustrative examples, FDA continues to address time criticality as a factor affecting both Criteria 3 and 4. Regardless of whether this seeming discrepancy was intentional or inadvertent, from the illustrative examples it remains clear that FDA will continue to consider time criticality as an important factor. For example, the guidance continues to list CDS tools used to detect possible stroke, sepsis, or myocardial infarction or to provide an alarm or alert to HCPs of possible patient deterioration as medical device software, “because they support time-critical decision making.” Similarly, FDA provides examples of CDS software that would fall outside the enforcement discretion policy when modified for a more time-critical use (e.g., software used to predict a cardiovascular event in the next 24 hours, to provide care pathway recommendations for acute back pain that may require immediate clinical intervention, or to predict in-hospital mortality to guide immediate escalation of care).
Less clear is how concerned FDA will remain about “automation bias.” As noted, automation bias is still addressed as a relevant factor in the guidance document, albeit now primarily in the context of Criterion 4 rather than Criterion 3. But in his CES remarks, Commissioner Makary seemed to reject the theory that FDA should “prevent the moral hazard of clinical staff turning their brain off and relying on whatever the computer or AI told them to do,” stating that “we need to be less paternalistic and instead trust clinicians to do what’s right for their patients.” How FDA applies this aspect of the revised guidance will merit close attention moving forward. Presumably, the emphasis will be more on time criticality, which is simpler to explain and understand than the more amorphous concept of automation bias.
Potential for Broader Digital Health and AI Reforms?
The Revised CDS Guidance and Revised General Wellness Guidance may be only the first step in a series of deregulatory actions by FDA to foster the development of AI-enabled digital health tools. In his remarks at the CES announcing the new guidances, Commissioner Makary said that FDA is working on “a new regulatory framework for AI . . . right now” that is “smarter and more forward-thinking.” A key goal of these regulatory actions would be to “modernize the agency and facilitate the AI revolution,” to ensure that the “US [will] be the best place for capital investment in AI.”
Commissioner Makary described a three-pronged approach to digital health reforms involving (1) deregulating low-risk products, (2) placing more emphasis on post-market monitoring, relying on big data, and (3) creating new guidances that anticipate “future AI products,” so that FDA is not merely in “reactionary mode” as new technologies are developed. At the same time, Commissioner Makary pointed to the large number of FDA guidance documents that deal with software and digital health, promising to “cut that number by about 50% or more and make them more clear, more concise, more modern, and more consistent.”
One area where Commissioner Makary signaled a potential change in FDA policy is AI-enabled image analysis software. Under current FDA regulations and the CDS guidance, even if medical image analysis software is not automated and only provides information for a doctor to consider, such software would be regulated by FDA as a medical device. In his speech at the CES, however, Commissioner Makary described an AI tool that “helps a doctor with colonoscopy images,” opining that “[i]f a doctor finds it useful . . . the doctor is ultimately responsible . . . [and] we’re not going to stop you from using it.” Conversely, “AI-generated information that is not purely informational, but is being used to trigger automatically medical interventions,” such as changing ventilator settings without physician involvement, was something that Commissioner Makary emphasized requires regulatory review and thorough evaluation. As with his comments on automation bias, these comments are difficult to reconcile with the current guidance and might signal that additional guidance will be forthcoming in the future. Alternatively, these comments might help reveal where FDA will focus its enforcement resources, even if the agency does not further clarify these aspects of the guidance.
Conclusion
The Revised CDS Guidance and Revised General Wellness Guidance are the most recent in a long line of new and revised guidance documents on digital health topics, which reflect the continuous evolution in FDA’s policies to address new learnings, technological advances, and shifting regulatory and political priorities.
Overall, the Revised CDS Guidance and Revised General Wellness Guidance represent a significant expansion of digital health technologies that FDA will consider as falling outside of active regulation as either CDS or a general wellness product. The changes address concerns raised by software developers and other stakeholders since those guidance documents were initially published, and they are in keeping with the current administration’s stated priorities to foster innovation and expand the availability of AI, wearable, and other digital health technologies.
It seems likely that more FDA guidance in related areas will be forthcoming as well. The CDRH guidance agenda for 2026 includes reissuing FDA’s 2022 “Policy for Device Software Functions” as draft guidance (currently this is final guidance, entitled “Policy for Device Software Functions and Mobile Medical Applications”). That guidance could be a vehicle for some of the policy changes to which Commissioner Makary alluded, including the new regulatory framework for AI that he has discussed. FDA may also make changes to existing guidance or publish new draft guidance documents that are not included in the CDRH agenda, as it did when it issued the Revised CDS Guidance and Revised General Wellness Guidance without prior notice.
It remains to be seen how the new policies will impact the digital health market. The practical implications of these and future digital health policy changes will depend on how the guidance is interpreted and applied by manufacturers and FDA in the real world.
Our team will continue to monitor developments in this area.
