That’s not what AI said: When AI gives the wrong answer in cosmetic dentistry

Jun 25, 2026, 10:32 by User Not Found
Anita Kemp, Case Manager at Dental Protection, shares a case study of a patient using AI to complain about cosmetic treatment.

When red flags whisper (and then scream) 

Dr G ran a busy inner-city cosmetic dental practice. Much of her work came via word-of-mouth referrals. After finding the clinic, patient Ms K was convinced Dr G was the right dentist for her to makeover her current teeth. She was excited, not just about improving her smile, but about beginning her transformation journey. 

What followed became a powerful lesson in clinical boundaries, patient expectations, and the importance of listening to professional intuition. 

The initial consultation: Excitement 

At their first appointment, the rapport was immediate. Ms K described feeling unhappy with the appearance of her teeth and how she was eager to make improvements. Dr G, experienced in cosmetic consultations, listened carefully, noted Ms K’s preferences, and reassured her that positive changes were achievable. 

Although Dr G personally felt Ms K’s natural smile was already acceptable, she respected her patient’s aesthetic concerns. A digital scan was taken and sent to the laboratory to create a veneer mock-up. When Ms K returned to review the proposed design, she was delighted. Both written and verbal consent were obtained to proceed. 

At this point, everything appeared straightforward. 

The first red flag 

When Ms K returned for tooth preparation two weeks later, something had changed. 

Reviewing the same mock-up, she insisted it was different from the one previously shown. She became visibly upset, tearful, and asked to call her husband. 

Dr G attempted to reassure her, replaying and reviewing the previous scans while gently confirming the design had not changed. She offered to place the temporary mock-up on so Ms K could “live with it” before committing to final preparation. 

Later, Ms K and her husband returned together. Both apologised for the earlier outburst. They explained that anxiety had likely influenced her reaction. They emphasised that she was not usually fussy and that she loved Dr G’s work. 

Despite an internal sense of hesitation, Dr G agreed to proceed. Minor adjustments were made, new scans were sent to the lab, and Ms K later emailed expressing satisfaction and excitement. 

That red flag moment passed. But it should not have. 

Treatment completion and temporary relief 

At the next appointment, the revised mock-up was reviewed. Both Ms K and her husband expressed satisfaction. Dr G once again clearly explained that this represented the best achievable outcome within her scope, and that changes would be difficult once final veneers were fabricated. 

They confirmed their approval. 

Minimal enamel preparation was completed, temporaries were placed, and final impressions sent to the laboratory. 

Dr G recalls feeling relief that the case appeared to be back on track. 

Four days later: Concern 

Four days after temporaries were cemented, emails began. 

Ms K had been searching online and comparing her own smile with that of strangers. She sent screenshots highlighting features she preferred in other patients’ results and circled aspects of her own temporaries she disliked. 

Notably, she also included: 

  • AI-generated smile simulations 
  • Written recommendations for clinical changes generated by AI 
  • Statements suggesting the requested modifications were “totally doable” 

At the review appointment, when Dr G tried to explain that the images she had sent through were not achievable in her situation, Ms K replied: “Well, that’s not what the AI said.” 

This marked a turning point. 

When expectations diverge from clinical reality 

Dr G reiterated the clinical limitations, reminding Ms K of earlier discussions regarding treatment and laboratory fabrication. She explained that treatment must be guided by clinical judgment and realistic anatomical parameters, not AI-generated suggestions based on photographs. 

At that moment, Dr G experienced what she later described as “that deep feeling of dread, when you realise you completely misread the situation”, recalling the earlier red flags she had noticed but chosen to ignore. 

  • The anxiety episode. 
  • The mock-up dispute. 
  • The insistence that something had changed when it had not. 

Dr G had initially considered specialist referral, but she had suppressed that instinct. Now the misalignment between expectation and reality was undeniable. 

The resolution 

Recognising that expectations were unlikely to align, Dr G recommended that Ms K stay in the temporary veneers and seek specialist opinion prior to proceeding to finish. She held firm this time, explaining that it was in everyone’s best interest that Ms K see a prosthodontist, and referral options were provided. 

Ms K declined referral, stating she would find another dentist herself. 

Ultimately, both parties agreed to a refund. Thankfully, only minimal enamel preparation had been performed, and full records were transferred to ensure continuity of care. 

The matter concluded without formal complaint, but not without reflection. 

Learning points 

This case highlights several important risk management themes: 

  • Early red flags matter. Emotional volatility, shifting perceptions, or disproportionate dissatisfaction, even if later minimised warrant careful consideration. 
  • Intuition is important. Clinical instinct is informed by experience. Discomfort is often an early indicator of misaligned expectations. 
  • Social media or online comparisons influence perception. Patients may compare themselves to curated images without understanding anatomical variability, case selection or digital enhancement. 
  • AI is not a treatment planner. AI-generated suggestions lack full clinical context, occlusal assessment, biological limitations and risk evaluation. Managing expectations around AI advice is becoming an emerging professional challenge. 
  • Referral is not failure. Referral can protect both patient and practitioner when expectations exceed what can reasonably be delivered. 
  • Documentation and clear warnings are critical. Dr G’s consistent reminders about limitations and documentation of these conversation and consent were essential in achieving an amicable resolution. 

Reflection 

Dr G later reflected that ego may have played a small role; wanting to satisfy the patient, wanting the case to succeed and wanting another positive outcome for her practice. 

But as is often the case, hindsight offered clarity: When your inner voice is signalling that something feels off, pause. Sometimes the most powerful risk management tool is the one we carry internally. Trust your intuition and your clinical judgment shaped by experience. 

Hindsight may be perfect insight. But foresight begins with listening.

The dentolegal journal from Dental Protection

Latest

That’s not what AI said: When AI gives the wrong answer in cosmetic dentistry

Jun 25, 2026, 10:32 by User Not Found
Anita Kemp, Case Manager at Dental Protection, shares a case study of a patient using AI to complain about cosmetic treatment.

When red flags whisper (and then scream) 

Dr G ran a busy inner-city cosmetic dental practice. Much of her work came via word-of-mouth referrals. After finding the clinic, patient Ms K was convinced Dr G was the right dentist for her to makeover her current teeth. She was excited, not just about improving her smile, but about beginning her transformation journey. 

What followed became a powerful lesson in clinical boundaries, patient expectations, and the importance of listening to professional intuition. 

The initial consultation: Excitement 

At their first appointment, the rapport was immediate. Ms K described feeling unhappy with the appearance of her teeth and how she was eager to make improvements. Dr G, experienced in cosmetic consultations, listened carefully, noted Ms K’s preferences, and reassured her that positive changes were achievable. 

Although Dr G personally felt Ms K’s natural smile was already acceptable, she respected her patient’s aesthetic concerns. A digital scan was taken and sent to the laboratory to create a veneer mock-up. When Ms K returned to review the proposed design, she was delighted. Both written and verbal consent were obtained to proceed. 

At this point, everything appeared straightforward. 

The first red flag 

When Ms K returned for tooth preparation two weeks later, something had changed. 

Reviewing the same mock-up, she insisted it was different from the one previously shown. She became visibly upset, tearful, and asked to call her husband. 

Dr G attempted to reassure her, replaying and reviewing the previous scans while gently confirming the design had not changed. She offered to place the temporary mock-up on so Ms K could “live with it” before committing to final preparation. 

Later, Ms K and her husband returned together. Both apologised for the earlier outburst. They explained that anxiety had likely influenced her reaction. They emphasised that she was not usually fussy and that she loved Dr G’s work. 

Despite an internal sense of hesitation, Dr G agreed to proceed. Minor adjustments were made, new scans were sent to the lab, and Ms K later emailed expressing satisfaction and excitement. 

That red flag moment passed. But it should not have. 

Treatment completion and temporary relief 

At the next appointment, the revised mock-up was reviewed. Both Ms K and her husband expressed satisfaction. Dr G once again clearly explained that this represented the best achievable outcome within her scope, and that changes would be difficult once final veneers were fabricated. 

They confirmed their approval. 

Minimal enamel preparation was completed, temporaries were placed, and final impressions sent to the laboratory. 

Dr G recalls feeling relief that the case appeared to be back on track. 

Four days later: Concern 

Four days after temporaries were cemented, emails began. 

Ms K had been searching online and comparing her own smile with that of strangers. She sent screenshots highlighting features she preferred in other patients’ results and circled aspects of her own temporaries she disliked. 

Notably, she also included: 

  • AI-generated smile simulations 
  • Written recommendations for clinical changes generated by AI 
  • Statements suggesting the requested modifications were “totally doable” 

At the review appointment, when Dr G tried to explain that the images she had sent through were not achievable in her situation, Ms K replied: “Well, that’s not what the AI said.” 

This marked a turning point. 

When expectations diverge from clinical reality 

Dr G reiterated the clinical limitations, reminding Ms K of earlier discussions regarding treatment and laboratory fabrication. She explained that treatment must be guided by clinical judgment and realistic anatomical parameters, not AI-generated suggestions based on photographs. 

At that moment, Dr G experienced what she later described as “that deep feeling of dread, when you realise you completely misread the situation”, recalling the earlier red flags she had noticed but chosen to ignore. 

  • The anxiety episode. 
  • The mock-up dispute. 
  • The insistence that something had changed when it had not. 

Dr G had initially considered specialist referral, but she had suppressed that instinct. Now the misalignment between expectation and reality was undeniable. 

The resolution 

Recognising that expectations were unlikely to align, Dr G recommended that Ms K stay in the temporary veneers and seek specialist opinion prior to proceeding to finish. She held firm this time, explaining that it was in everyone’s best interest that Ms K see a prosthodontist, and referral options were provided. 

Ms K declined referral, stating she would find another dentist herself. 

Ultimately, both parties agreed to a refund. Thankfully, only minimal enamel preparation had been performed, and full records were transferred to ensure continuity of care. 

The matter concluded without formal complaint, but not without reflection. 

Learning points 

This case highlights several important risk management themes: 

  • Early red flags matter. Emotional volatility, shifting perceptions, or disproportionate dissatisfaction, even if later minimised warrant careful consideration. 
  • Intuition is important. Clinical instinct is informed by experience. Discomfort is often an early indicator of misaligned expectations. 
  • Social media or online comparisons influence perception. Patients may compare themselves to curated images without understanding anatomical variability, case selection or digital enhancement. 
  • AI is not a treatment planner. AI-generated suggestions lack full clinical context, occlusal assessment, biological limitations and risk evaluation. Managing expectations around AI advice is becoming an emerging professional challenge. 
  • Referral is not failure. Referral can protect both patient and practitioner when expectations exceed what can reasonably be delivered. 
  • Documentation and clear warnings are critical. Dr G’s consistent reminders about limitations and documentation of these conversation and consent were essential in achieving an amicable resolution. 

Reflection 

Dr G later reflected that ego may have played a small role; wanting to satisfy the patient, wanting the case to succeed and wanting another positive outcome for her practice. 

But as is often the case, hindsight offered clarity: When your inner voice is signalling that something feels off, pause. Sometimes the most powerful risk management tool is the one we carry internally. Trust your intuition and your clinical judgment shaped by experience. 

Hindsight may be perfect insight. But foresight begins with listening.

Global news

That’s not what AI said: When AI gives the wrong answer in cosmetic dentistry

Jun 25, 2026, 10:32 by User Not Found
Anita Kemp, Case Manager at Dental Protection, shares a case study of a patient using AI to complain about cosmetic treatment.

When red flags whisper (and then scream) 

Dr G ran a busy inner-city cosmetic dental practice. Much of her work came via word-of-mouth referrals. After finding the clinic, patient Ms K was convinced Dr G was the right dentist for her to makeover her current teeth. She was excited, not just about improving her smile, but about beginning her transformation journey. 

What followed became a powerful lesson in clinical boundaries, patient expectations, and the importance of listening to professional intuition. 

The initial consultation: Excitement 

At their first appointment, the rapport was immediate. Ms K described feeling unhappy with the appearance of her teeth and how she was eager to make improvements. Dr G, experienced in cosmetic consultations, listened carefully, noted Ms K’s preferences, and reassured her that positive changes were achievable. 

Although Dr G personally felt Ms K’s natural smile was already acceptable, she respected her patient’s aesthetic concerns. A digital scan was taken and sent to the laboratory to create a veneer mock-up. When Ms K returned to review the proposed design, she was delighted. Both written and verbal consent were obtained to proceed. 

At this point, everything appeared straightforward. 

The first red flag 

When Ms K returned for tooth preparation two weeks later, something had changed. 

Reviewing the same mock-up, she insisted it was different from the one previously shown. She became visibly upset, tearful, and asked to call her husband. 

Dr G attempted to reassure her, replaying and reviewing the previous scans while gently confirming the design had not changed. She offered to place the temporary mock-up on so Ms K could “live with it” before committing to final preparation. 

Later, Ms K and her husband returned together. Both apologised for the earlier outburst. They explained that anxiety had likely influenced her reaction. They emphasised that she was not usually fussy and that she loved Dr G’s work. 

Despite an internal sense of hesitation, Dr G agreed to proceed. Minor adjustments were made, new scans were sent to the lab, and Ms K later emailed expressing satisfaction and excitement. 

That red flag moment passed. But it should not have. 

Treatment completion and temporary relief 

At the next appointment, the revised mock-up was reviewed. Both Ms K and her husband expressed satisfaction. Dr G once again clearly explained that this represented the best achievable outcome within her scope, and that changes would be difficult once final veneers were fabricated. 

They confirmed their approval. 

Minimal enamel preparation was completed, temporaries were placed, and final impressions sent to the laboratory. 

Dr G recalls feeling relief that the case appeared to be back on track. 

Four days later: Concern 

Four days after temporaries were cemented, emails began. 

Ms K had been searching online and comparing her own smile with that of strangers. She sent screenshots highlighting features she preferred in other patients’ results and circled aspects of her own temporaries she disliked. 

Notably, she also included: 

  • AI-generated smile simulations 
  • Written recommendations for clinical changes generated by AI 
  • Statements suggesting the requested modifications were “totally doable” 

At the review appointment, when Dr G tried to explain that the images she had sent through were not achievable in her situation, Ms K replied: “Well, that’s not what the AI said.” 

This marked a turning point. 

When expectations diverge from clinical reality 

Dr G reiterated the clinical limitations, reminding Ms K of earlier discussions regarding treatment and laboratory fabrication. She explained that treatment must be guided by clinical judgment and realistic anatomical parameters, not AI-generated suggestions based on photographs. 

At that moment, Dr G experienced what she later described as “that deep feeling of dread, when you realise you completely misread the situation”, recalling the earlier red flags she had noticed but chosen to ignore. 

  • The anxiety episode. 
  • The mock-up dispute. 
  • The insistence that something had changed when it had not. 

Dr G had initially considered specialist referral, but she had suppressed that instinct. Now the misalignment between expectation and reality was undeniable. 

The resolution 

Recognising that expectations were unlikely to align, Dr G recommended that Ms K stay in the temporary veneers and seek specialist opinion prior to proceeding to finish. She held firm this time, explaining that it was in everyone’s best interest that Ms K see a prosthodontist, and referral options were provided. 

Ms K declined referral, stating she would find another dentist herself. 

Ultimately, both parties agreed to a refund. Thankfully, only minimal enamel preparation had been performed, and full records were transferred to ensure continuity of care. 

The matter concluded without formal complaint, but not without reflection. 

Learning points 

This case highlights several important risk management themes: 

  • Early red flags matter. Emotional volatility, shifting perceptions, or disproportionate dissatisfaction, even if later minimised warrant careful consideration. 
  • Intuition is important. Clinical instinct is informed by experience. Discomfort is often an early indicator of misaligned expectations. 
  • Social media or online comparisons influence perception. Patients may compare themselves to curated images without understanding anatomical variability, case selection or digital enhancement. 
  • AI is not a treatment planner. AI-generated suggestions lack full clinical context, occlusal assessment, biological limitations and risk evaluation. Managing expectations around AI advice is becoming an emerging professional challenge. 
  • Referral is not failure. Referral can protect both patient and practitioner when expectations exceed what can reasonably be delivered. 
  • Documentation and clear warnings are critical. Dr G’s consistent reminders about limitations and documentation of these conversation and consent were essential in achieving an amicable resolution. 

Reflection 

Dr G later reflected that ego may have played a small role; wanting to satisfy the patient, wanting the case to succeed and wanting another positive outcome for her practice. 

But as is often the case, hindsight offered clarity: When your inner voice is signalling that something feels off, pause. Sometimes the most powerful risk management tool is the one we carry internally. Trust your intuition and your clinical judgment shaped by experience. 

Hindsight may be perfect insight. But foresight begins with listening.

Sed ut perspiciatis unde omnis iste natus error sit voluptatem accusantium
CTA