WASHINGTON — The same artificial intelligence that helps doctors spot tumors faster is now cheap enough and simple enough for anyone to misuse. That is the central warning from a new SafeWise analysis and from experts who track fraud in American healthcare. The barrier has collapsed.
Dr. Musheer Ahmed, who runs the AI fraud-detection firm Codoxo, told SafeWise that powerful image and video generation tools are now widely available. The record shows that the technology driving genuine productivity gains in medicine has been turned into a weapon.
Two distinct threats
The SafeWise report draws a clear line between two kinds of medical deepfakes. On one side are manipulated diagnostic images — X-rays, CT scans, MRIs — altered to show a disease that is not there or to hide one that is. On the other side are fabricated videos of doctors or public health officials giving false medical advice.
Both categories, the report argues, do the same thing: erode trust in medical institutions and cause real harm. A faked radiology image can lead directly to a misdiagnosis.
It can produce a wrong treatment plan. It can delay necessary care. A deepfake video of a trusted physician endorsing a bogus treatment could send thousands of patients down a dangerous path.
The tools needed to generate either are now online, often free. Ahmed’s firm operates on the front lines of this problem.
Codoxo uses AI to find anomalies in healthcare claims and billing data. The irony is obvious to industry observers: the same class of technology defending the system is the one attacking it.
How we got here
Generative AI has advanced at a speed that regulators and healthcare institutions have not matched. The SafeWise analysis is a public reckoning with that gap. For years, the conversation about AI in medicine centered on diagnosis, drug discovery, and administrative efficiency.
That conversation has shifted. The record shows that the same image-generation models that can create a realistic medical illustration can also alter a real one.
The same voice-cloning software that can help a patient who has lost the ability to speak can also impersonate a doctor. The technology is not inherently malicious. But the access to it is now universal, and the safeguards are not.
Ahmed made the point bluntly. “Today, there are very strong image and video generation capabilities,” he said.
The statement is a factual description of a problem healthcare is only beginning to confront.
Trust as the casualty
The SafeWise report is not just about technical risk. It is about what happens when patients cannot trust what they see. A doctor’s reputation, built over years, can be destroyed in seconds by a convincing fake.
A public health message, carefully crafted, can be undermined by a fabricated alternative that spreads faster. That erosion of trust is the deeper public health risk.
It does not require a single successful attack. It only requires the possibility of one. Once people know that medical images can be faked, they may question real ones.
Once they know that a video of a doctor can be fabricated, they may dismiss authentic guidance. The healthcare industry is now in a race.
Defenders like Codoxo are building AI to catch the fakes. But the attackers have the same tools and fewer constraints. The SafeWise analysis makes clear that the threat is not theoretical.
It is present. It is growing. And it is only beginning to be confronted.
{“@context”: “https://schema.org”, “@type”: “NewsArticle”, “headline”: “AI Tools That Aid Diagnosis Now Cheap Enough for Widespread Misuse”, “description”: “WASHINGTON — The same artificial intelligence that helps doctors spot tumors faster is now cheap enough and simple enough for anyone to misuse. That is…”, “datePublished”: “2026-06-25T09:12:00+00:00T09:30:00+00:00”, “author”: {“@type”: “Person”, “name”: “Sebastian Smith”}, “publisher”: {“@type”: “Organization”, “name”: “InfoPulse Today”, “url”: “https://newsfeed.infopulsetoday.com/”}}
