![\"artificial-intelligence-3706562_1280\"](\"https:\/\/escope.ages.com.au\/july-2025\/wp-content\/uploads\/sites\/10\/2025\/05\/artificial-intelligence-3706562_1280.jpg\" "\\"artificial-intelligence-3706562_1280\\"")
\n\n\u00a0Author: Tracy Pickett, Avant lawyer and policy adviser and Dr Vanessa Watson, Obstetrician and Gynaecologist, Sunshine Coast, QLD.<\/p>\n
Editor and Interviewer: Dr Rebecca Nash.<\/strong>\u00a0<\/p>\n In just six months, nearly one in five Avant members surveyed (Haysom, 2025) said they had already used an AI scribe, and nearly half were considering it. As the uptake of artificial intelligence in clinical documentation rapidly accelerates, many doctors are left wondering: how do we balance efficiency with legal and ethical responsibility?<\/p>\n To explore the medico-legal implications of using AI in private practice particularly for documentation, we sat down with Tracy Pickett, Avant lawyer and policy adviser. She unpacks the core issues doctors should consider in relation to using AI in practice: consent, privacy, accuracy, and the importance of having clear policies and procedures.<\/p>\n With growing interest in AI scribes, where should doctors begin when considering the medico-legal risks<\/b>?<\/b><\/p>\n Obtaining consent is crucial. We are often asked why it’s necessary when the AI is only helping to generate patient records, but there are three key reasons:<\/p>\n Beyond consent, privacy remains a key area of concern. What are the main risks practices should watch for?<\/b><\/p>\n There are a few. First, these AI tools may store or process data overseas, which triggers obligations under the Privacy Act. You need to know exactly where your patients’ data is going and whether the AI scribe meets Australian privacy standards. Secondly, even if data is stored locally, integration with practice management systems can create weak points for potential data breaches. Doctors have a legal duty to protect patient confidentiality, and that extends to digital systems you choose to bring into your practice. This means there is an even greater need to tell patients that AI is being used so they understand its interaction with their information.<\/p>\n Accuracy is another concern. How reliable are AI-generated notes in practice?<\/b><\/p>\n You can’t assume the output is always correct. We’ve seen cases where the AI missed key clinical details or misinterpreted what was said. Ultimately, the doctor is responsible for the final record. If a claim or complaint arises, it’s the doctor’s name on the record, not the AI. So yes, AI can improve efficiency, but it can also introduce risk if not reviewed properly. The legal expectation is still that the record reflects an accurate and complete account of the consultation, regardless of how it is prepared.<\/p>\n Having the right safeguards matters. What kinds of policies and procedures should practices establish?<\/b><\/p>\n At a minimum, practices should have clear policies around the use of AI tools. This includes:<\/p>\n Staff training is essential too. If multiple clinicians are using the system, there should be consistency in application of the tools. \u00a0You also need to document these processes, because if a complaint is lodged, you’ll need to show what protocols were in place to safeguard patient care and privacy.<\/p>\n Before we wrap up, any final advice for doctors thinking about adopting these tools?<\/b><\/p>\n Don’t be afraid of innovation, but don’t assume the legal risks disappear just because something is marketed as “smart” or “secure” and is easy to use. Take the time to understand how the tool works, and make sure it fits within your broader medico-legal responsibilities. Seek clarification or advice if you need to. AI can be a great support, but it doesn’t replace your clinical judgement or your legal obligations.<\/p>\n If you’re looking for more detailed guidance, I’d recommend exploring our dedicated AI resources on the Avant AI collection <\/a>page. We’ve brought together articles, case studies and practical checklists to help you introduce these tools safely and legally.<\/p>\n We now ask Dr Vanessa Watson, Obstetrician and Gynaecologist on the Sunshine Coast, Queensland about her experience using an AI scribe as she has just introduced it into her private practice.<\/em><\/p>\n How has the integration of AI for medical documentation affected your daily workflow, particularly in terms of efficiency, accuracy, and time management during patient consultations?”<\/strong><\/p>\n Using medical AI has saved me 30-60 minutes per full day of consulting by streamlining documentation and GP letters. This helps me see more patients or finish on time-especially when appointments involve lengthy histories or counselling. I typically make only 3-5 minor edits per patient, and while occasional discrepancies need close checking, review is quick. I haven’t yet customised formatting, but the default setup works well for now.<\/p>\n At the moment do you gain verbal or written consent from patients and so far has anyone declined for you to use AI?<\/strong><\/p>\n At our practice we gain consent in two ways for each patient: written information and written consent on arrival to the practice and verbal consent at the start of the consultation. I also provide information on the use of AI on my website. I have had one patient decline so far.<\/p>\n “Overall, how do you feel about using AI in your documentation process-has it made your job more enjoyable or challenging?”<\/strong><\/p>\n Overall, the use of AI has returned a lot of enjoyment to my job. It feels like I can focus on being a doctor and building a relationship with patients, rather than getting bogged down with paperwork. I am looking forward to being more creative in the ways I can use it to further enrich my practice.<\/p>\n References<\/p>\n Haysom, G. (2025). AI scribes: Rapid uptake in 6 months<\/em>. Avant. https:\/\/avant.org.au\/resources\/ai-scribes-rapid-uptake-in-6-months<\/a><\/p>\n Office of the Australian Information Commissioner. (2024). Guidance on privacy and the use of commercially available AI products<\/em>. https:\/\/www.oaic.gov.au\/privacy\/privacy-guidance-for-organisations-and-government-agencies\/guidance-on-privacy-and-the-use-of-commercially-available-ai-products<\/a><\/p>\n Privacy Act 1988<\/em> (Cth). https:\/\/www.legislation.gov.au\/C2004A03712<\/a><\/p>\n\t\t\t Obtaining consent is crucial. We are often asked why it’s necessary when the AI is only helping to generate patient records, but there are three key reasons:<\/p>\n There are a few. First, these AI tools may store or process data overseas, which triggers obligations under the Privacy Act. You need to know exactly where your patients’ data is going and whether the AI scribe meets Australian privacy standards. Secondly, even if data is stored locally, integration with practice management systems can create weak points for potential data breaches. Doctors have a legal duty to protect patient confidentiality, and that extends to digital systems you choose to bring into your practice. This means there is an even greater need to tell patients that AI is being used so they understand its interaction with their information.<\/p>\n\t\t\t You can’t assume the output is always correct. We’ve seen cases where the AI missed key clinical details or misinterpreted what was said. Ultimately, the doctor is responsible for the final record. If a claim or complaint arises, it’s the doctor’s name on the record, not the AI. So yes, AI can improve efficiency, but it can also introduce risk if not reviewed properly. The legal expectation is still that the record reflects an accurate and complete account of the consultation, regardless of how it is prepared.<\/p>\n\t\t\t At a minimum, practices should have clear policies around the use of AI tools. This includes:<\/p>\n Staff training is essential too. If multiple clinicians are using the system, there should be consistency in application of the tools. \u00a0You also need to document these processes, because if a complaint is lodged, you’ll need to show what protocols were in place to safeguard patient care and privacy.<\/p>\n\t\t\t Don’t be afraid of innovation, but don’t assume the legal risks disappear just because something is marketed as “smart” or “secure” and is easy to use. Take the time to understand how the tool works, and make sure it fits within your broader medico-legal responsibilities. Seek clarification or advice if you need to. AI can be a great support, but it doesn’t replace your clinical judgement or your legal obligations.<\/p>\n If you’re looking for more detailed guidance, I’d recommend exploring our dedicated AI resources on the Avant AI collection <\/a>page. We’ve brought together articles, case studies and practical checklists to help you introduce these tools safely and legally.<\/p>\n We now ask Dr Vanessa Watson, Obstetrician and Gynaecologist on the Sunshine Coast, Queensland about her experience using an AI scribe as she has just introduced it into her private practice.<\/em><\/p>\n\t\t\t Using medical AI has saved me 30-60 minutes per full day of consulting by streamlining documentation and GP letters. This helps me see more patients or finish on time-especially when appointments involve lengthy histories or counselling. I typically make only 3-5 minor edits per patient, and while occasional discrepancies need close checking, review is quick. I haven’t yet customised formatting, but the default setup works well for now.<\/p>\n\t\t\t At our practice we gain consent in two ways for each patient: written information and written consent on arrival to the practice and verbal consent at the start of the consultation. I also provide information on the use of AI on my website. I have had one patient decline so far.<\/p>\n\t\t\t Overall, the use of AI has returned a lot of enjoyment to my job. It feels like I can focus on being a doctor and building a relationship with patients, rather than getting bogged down with paperwork. I am looking forward to being more creative in the ways I can use it to further enrich my practice.<\/p>\n\t\t\t Haysom, G. (2025). AI scribes: Rapid uptake in 6 months<\/em>. Avant. https:\/\/avant.org.au\/resources\/ai-scribes-rapid-uptake-in-6-months<\/a><\/p>\n Office of the Australian Information Commissioner. (2024). Guidance on privacy and the use of commercially available AI products<\/em>. https:\/\/www.oaic.gov.au\/privacy\/privacy-guidance-for-organisations-and-government-agencies\/guidance-on-privacy-and-the-use-of-commercially-available-ai-products<\/a><\/p>\n\n
\n
With growing interest in AI scribes, where should doctors begin when considering the medico-legal risks?<\/h3>\t\t\t\n\t\t\t\t
\n
Beyond consent, privacy remains a key area of concern. What are the main risks practices should watch for?<\/h3>\t\t\t\n\t\t\t\t
Accuracy is another concern. How reliable are AI-generated notes in practice?<\/h3>\t\t\t\n\t\t\t\t
Having the right safeguards matters. What kinds of policies and procedures should practices establish?<\/h3>\t\t\t\n\t\t\t\t
\n
Before we wrap up, any final advice for doctors thinking about adopting these tools?<\/h3>\t\t\t\n\t\t\t\t
How has the integration of AI for medical documentation affected your daily workflow, particularly in terms of efficiency, accuracy, and time management during patient consultations?<\/h3>\t\t\t\n\t\t\t\t
At the moment do you gain verbal or written consent from patients and so far has anyone declined for you to use AI?<\/h3>\t\t\t\n\t\t\t\t
Overall, how do you feel about using AI in your documentation process-has it made your job more enjoyable or challenging?<\/h3>\t\t\t\n\t\t\t\t
References<\/h3>\t\t\t\n\t\t\t\t
![\"Dr](\"https:\/\/escope.ages.com.au\/july-2025\/wp-content\/uploads\/sites\/10\/2025\/05\/escope-author__dr-tracey-pickett-001.avif\" "\\"Dr")
\n\t\t\t\t![\"Dr](\"https:\/\/escope.ages.com.au\/july-2025\/wp-content\/uploads\/sites\/10\/2025\/05\/escope-author__dr-rebecca-nash-001.avif\" "\\"Dr")
\n\t\t\t\t![\"Dr](\"https:\/\/escope.ages.com.au\/july-2025\/wp-content\/uploads\/sites\/10\/2025\/05\/escope-author__dr-vanessa-watson-001.avif\" "\\"Dr")
\n\t\t\t\t