Google's AI Overview gives incorrect diabetes blood sugar levels, sourcing advice from major Indian hospitals. For instance, it recommends a 180 mg/dl post-meal sugar threshold, despite Indian medical guidelines setting it at less than 200 mg/dl. This subtle, 20-point discrepancy puts patients at risk because it feels correct coming from a reputable source.
How We Got Here
Corporate hospitals like Apollo and Max have long optimized their websites with reams of medical content for Google's search algorithms. By June, a healthcare content consultant audited nearly 500 articles from five top chains, revealing widespread use of AI-generated content.
The Numbers
- The audit covered five major corporate hospital chains: Apollo, Max, Fortis, Medanta, and Artemis.
- It found articles promoting incorrect wait times for procedures and recommending discontinued drugs.
- Content often used Western clinical benchmarks for Indian patients and outdated emergency protocols.
- Obvious AI "tells" included mistranslating "menstrual discharge" to "sewage water" and "May" (month) to "can" (verb).
What Happens Next
🇮🇳 Why This Matters for India
For the millions of healthcare consumers in Tier-2 and Tier-3 cities relying on online search for medical advice, this loop risks real harm from subtle misinformation on critical health conditions.
The Take
This reveals a stark content governance and trust crisis, amplified by brand authority, not a mere AI hallucination problem. Hospitals are implicitly offloading their medical liability to unverified AI, prioritizing search engine rankings over patient safety.
Source:
The Ken ↗