In a series of posts on Twitter, Elon Musk promoted the site’s “Grok” AI chatbot as being capable of analyzing blood test results, X-rays, and other medical information, inviting the site’s users to try the function. Musk also went on to state that, in the future, the bot will also be able to analyze legal documents. These claims may put Twitter at risk of violating laws against the unlicensed practice of medicine, as well as the unlicensed practice of law. 

Bottom Line: Elon Musk’s push to commercialize Twitter’s “Grok” AI model and salvage the platform’s profitability is taking him headlong into uncharted legal territory, with potentially serious consequences. 

 

Elon Musk Promotes “Grok” AI Model as Being Capable of Analyzing Medical Data, Legal Documents

On October 29, Elon Musk made a post on Twitter in which he advertised his company’s AI chatbot “Grok” as being capable of reading PET, MRI, and x-ray scans, as well as other medical information:

“Try submitting x-ray, PET, MRI or other medical images to Grok for analysis. This is still early stage [sic], but it is already quite accurate and will become extremely good. Let us know where Grok gets it right or needs work.”

When asked if the bot could also analyze blood sample results, Musk stated that it could. and also said that the ability to read and analyze legal documents was forthcoming. At the time of this writing, Musk’s post has received at least 41 million views, and the replies seem to suggest that many users did, in fact, submit medical information to Grok for analysis, despite a lack of transparency on how this data might be used.

Possible Unlicensed Practice of Medicine

At present, Twitter is in the process of relocating from its former San Francisco, CA headquarters to Austin, TX. Under Texas law, it is likely that Grok is engaging in the provision of medical advice, and is thus unlawfully practicing medicine without a license. The practice of medicine is defined by Texas law as: “the diagnosis, treatment, or offer to treat a mental or physical disease or disorder or a physical deformity or injury by any system or method, or the attempt to effect cures of those conditions, by a person who:

1. publicly professes to be a physician or surgeon; or
2. directly or indirectly charges money or other compensation for those services.” (Tex. Occ. Code § 151.002(13).)

Although Musk has held out Grok as being competent to interpret medical information, he has made no claims that Grok is a licensed doctor. However, access to Grok is a feature exclusive to paid Twitter users, meaning that money is being (indirectly) charged for the provision of services. This makes Grok’s services, by definition, paid.

Grok’s interpretation of medical scans and results counts as the practice of medicine. In Lowry v. Texas Medical Board (2021), the Texas 3rd Court of Appeals found that a former physician who was being consulted with by a licensed physician to “interpret” patients’ EEG results was, indeed, practicing medicine within the meaning of Texas state statute, even though he had no direct contact with patients. (Lowry v. Texas Medical Board, 2021 Tex. App. LEXIS 4656, at *13-14.)

Grok’s actions also can be easily construed as the corporate practice of medicine. In fact, it is quite difficult to construe them as anything else. Similar to the legal profession’s ban on non-lawyers having ownership interests in law firms, Texas law explicitly prohibits the corporate practice of medicine, on grounds that doing so would potentially influence the professional judgment of doctors. (Tex. Occ. Code Ann. 164.052(a)(17)) In Gupta v. Eastern Idaho Tumor Institute, the 14th District held that a joint venture between an oncologist and a corporation to operate a cancer treatment center was not in violation of state law because the arrangement posed no risk of affecting the doctor’s professional judgment. (140 S.W.3d 747 (Tex. App. Houston 14th 2004)) However, as an AI construct, Grok’s fundamental “judgment” (if such a characteristic can truly be ascribed to an AI) is quite literally in the hands of Twitter, a corporate entity not licensed to practice medicine. If we interpret Grok as “practicing medicine,” this is facially illegal.

(It should be noted that while some have read Texas statute to permit the practice of medicine by LLCs, there are a number of instances of LLCs being prosecuted for doing precisely this.)

Possible (Future) Unauthorized Practice of Law

Advertising Grok as being capable of reading legal documents also carries significant risk of constituting the unauthorized practice of law. Though it remains to be seen as to exactly how Musk plans to implement the “legal analysis” feature into the AI model, if its actions rose to the level of constituting the practice of law, it would likely run afoul of Rule 5.4(d)(3) of the ABA Model Rules of Professional Conduct, for the same reasons Grok’s current actions potentially constitute the corporate practice of medicine. 5.4(d) (3) forbids legal practice in any form where a non-lawyer has control over the professional judgment of a lawyer. Though Grok is obviously not a lawyer, such an arrangement would clearly go against the intent of the law.

 

Matthew Sparks was a Justice Fellow at the Tech Institute 2023-2024.