{"id":46315,"topic":"ai","source":"Wealth Briefing","title":"OPINION OF THE WEEK: Beware Of AI Notetaking Threat To Confidentiality - Wealth Briefing","url":"https://www.wealthbriefing.com/html/article.php/opinion-of-the-week%3A-beware-of-ai-notetaking-threat-to-confidentiality-","url_hash":"c21b5b9340fa1d9b7d0b653309dd170bb3188a23","author":"","summary":"OPINION OF THE WEEK: Beware Of AI Notetaking Threat To Confidentiality Wealth Briefing","content":"Lawyers can put confidentiality and privilege at risk if AI notetakers are used in meetings – one of the risks of this burgeoning technology that HNW individuals need to understand.\n In thinking about all the use cases and risks that attend AI in\n today’s private client advisory sector, one term that ought to\n stick in the mind right now is “confidentiality.”\n \n For those who don’t follow the twists and turns of AI closely, as\n well as legal cases, it might not appear obvious that one of the\n dangers of AI if not used properly is the destruction of\n lawyer-client privilege. In an industry where HNW and UHNW\n individuals and families understandably value their privacy, this\n is a big problem. \n \n Many of us who go on Zoom, Teams and other platforms for business\n meetings might be told in advance – as we should – that “this\n meeting is being recorded” – usually in a cheerful voice. For\n journalists who haven’t mastered the grinding skill of shorthand\n – as I did decades ago – all these recording tools are a\n godsend. \n \n But there is a downside.\n \n As noted by the law firm White & Case in an April 23 report,\n entitled Attorney-client privilege and work product in the\n age of generative AI, two cases illustrate the risks –\n with very different results. \n \n On 10 February this year, the US District Court for the Eastern\n District of Michigan in Warner v. Gilbarco,\n Inc denied a motion to compel production of documents\n that a self-represented litigant had prepared using a public AI\n chatbot. The court found that work product protection\n applied. \n \n A week later, on 17 February, the US District Court for the\n Southern District of New York in United States vs.\n Heppner reached the opposite result, ordering a criminal\n defendant to produce documents he had generated using another\n public AI chatbot while seeking legal advice. \n \n The Heppner case was also mentioned at a recent media webinar,\n which I attended. The event was hosted by law firm ArentFox Schiff.\n Sarah Severson, a partner, explained the risks of AI in meetings\n where there are not clear understandings about what is\n involved.\n \n Notes taken via AI could be “discoverable,” she continued.\n Automated transcripts and summaries are classed as digital\n documents that opposing counsel can subpoena in civil or criminal\n cases. \n \n Severson said lawyers should always disclose their use of AI to\n clients in their engagement letters. \n \n “AI does not owe clients a duty of confidentiality or\n accountability,” she said. \n \n The root of the problem is that cloud-based processing,\n third-party vendor data access, and terms of service permitting\n model training eliminate the legal expectation of\n confidentiality. \n \n This is a global issue.\n \n The UK is an example, as recent cases demonstrate. In the case of\n UK v Secretary of State for the Home Department [2026] UKUT\n 81 (Hamid), the Upper Tribunal (Immigration and Asylum\n Chamber) delivered the first decision by an English court or\n tribunal to directly address what the legal professional\n privilege risks when confidential and privileged material is\n uploaded to open-source AI tools. \n \n It is important to remember that jurisdictions such as\n Singapore share the common law traditions of the UK and US, for\n example. As case law builds up around the world, the AI\n notetaking issue will become a global one. \n \n One of the points coming out of all this is that – as Severson\n said in the webinar – AI is a valuable tool. The ability to\n summarize and collect data is useful, to give just one\n example. The arrival of AI is, it should be said, also a\n reason why the grunt work that junior lawyers used to do is being\n replaced. These recent cases are also a reminder that large\n AI models, which rely on vast amounts of data, are, in a\n way, a sort of “public” field. \n \n Fintechs are starting to address the problem of how public AI can\n be. In\n early June, Custodia, a Swiss privacy-first AI startup, said\n it had launched Sentinel, a “physical AI thinking appliance\n developed and designed from the ground up for executives, family\n offices, scientific researchers, and any professional whose\n intellectual property is too valuable and too sensitive to trust\n to the cloud.”\n \n Perhaps it is not a coincidence that Custodia is Swiss – the land\n of bank secrecy (albeit no longer on cross-border matters),\n where privacy is still highly prized, as it should\n be. \n \n In its press release about the launch, Custodia said: “Rather\n than relying on pre-trained knowledge, Sentinel ingests your\n documents, understands their context, and retrieves precisely the\n right information to ground every answer it gives. Load thousands\n of files – financial records, research papers, legal\n documents, corporate history, correspondence – and Sentinel\n draws only from that store of knowledge. No hallucination from\n unrelated internet data.”\n \n Well, that is the sales pitch, and this news service is looking\n into this area about private AI to see just how thick the walls\n of privacy really are. What is clear, however, is that if\n you are a lawyer or a client, or indeed a professional wealth\n manager, investor or professional figure having a confidential\n discussion, the AI recording gizmo should be turned off.\n Or, at the very least, the use of these devices must be\n clearly disclosed ahead of time, giving affected parties a chance\n to refuse.","image_url":null,"lang":"en","published_at":"2026-07-26T07:07:10+00:00","fetched_at":"2026-07-26T07:15:05+00:00","status":"read","starred":0,"extract_state":"ok","summary_auto":"Lawyers can put confidentiality and privilege at risk if AI notetakers are used in meetings – one of the risks of this burgeoning technology that HNW individuals need to understand. In thinking about all the use cases and risks that attend AI in\n today’s private client advisory sector, one term that ought to\n stick in the mind right now is “confidentiality.”\n \n For those who don’t follow the twists and turns of AI closely, as\n well as legal cases, it might not appear obvious that one of the\n dangers of AI if not used properly is the destruction of\n lawyer-client privilege.","cluster_id":null,"extract_retries":0,"extract_error":null,"contract_version":"news_item.v1","format_contract_version":"news_item_formats.v1","dedup_url":"https://www.wealthbriefing.com/html/article.php/opinion-of-the-week%3A-beware-of-ai-notetaking-threat-to-confidentiality-","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 5514 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":5514,"summary_length":586,"usable_text_length":5514,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":5514,"summary_length":586}},"news_item":{"id":46315,"canonical_url":"https://www.wealthbriefing.com/html/article.php/opinion-of-the-week%3A-beware-of-ai-notetaking-threat-to-confidentiality-","source_url":"https://www.wealthbriefing.com/html/article.php/opinion-of-the-week%3A-beware-of-ai-notetaking-threat-to-confidentiality-","title":"OPINION OF THE WEEK: Beware Of AI Notetaking Threat To Confidentiality - Wealth Briefing","source_name":"Wealth Briefing","author":null,"published_at":"2026-07-26T07:07:10+00:00","locale":"en","topic":"ai","tags":[],"rss_summary":"OPINION OF THE WEEK: Beware Of AI Notetaking Threat To Confidentiality Wealth Briefing","full_text":"Lawyers can put confidentiality and privilege at risk if AI notetakers are used in meetings – one of the risks of this burgeoning technology that HNW individuals need to understand.\n In thinking about all the use cases and risks that attend AI in\n today’s private client advisory sector, one term that ought to\n stick in the mind right now is “confidentiality.”\n \n For those who don’t follow the twists and turns of AI closely, as\n well as legal cases, it might not appear obvious that one of the\n dangers of AI if not used properly is the destruction of\n lawyer-client privilege. In an industry where HNW and UHNW\n individuals and families understandably value their privacy, this\n is a big problem. \n \n Many of us who go on Zoom, Teams and other platforms for business\n meetings might be told in advance – as we should – that “this\n meeting is being recorded” – usually in a cheerful voice. For\n journalists who haven’t mastered the grinding skill of shorthand\n – as I did decades ago – all these recording tools are a\n godsend. \n \n But there is a downside.\n \n As noted by the law firm White & Case in an April 23 report,\n entitled Attorney-client privilege and work product in the\n age of generative AI, two cases illustrate the risks –\n with very different results. \n \n On 10 February this year, the US District Court for the Eastern\n District of Michigan in Warner v. Gilbarco,\n Inc denied a motion to compel production of documents\n that a self-represented litigant had prepared using a public AI\n chatbot. The court found that work product protection\n applied. \n \n A week later, on 17 February, the US District Court for the\n Southern District of New York in United States vs.\n Heppner reached the opposite result, ordering a criminal\n defendant to produce documents he had generated using another\n public AI chatbot while seeking legal advice. \n \n The Heppner case was also mentioned at a recent media webinar,\n which I attended. The event was hosted by law firm ArentFox Schiff.\n Sarah Severson, a partner, explained the risks of AI in meetings\n where there are not clear understandings about what is\n involved.\n \n Notes taken via AI could be “discoverable,” she continued.\n Automated transcripts and summaries are classed as digital\n documents that opposing counsel can subpoena in civil or criminal\n cases. \n \n Severson said lawyers should always disclose their use of AI to\n clients in their engagement letters. \n \n “AI does not owe clients a duty of confidentiality or\n accountability,” she said. \n \n The root of the problem is that cloud-based processing,\n third-party vendor data access, and terms of service permitting\n model training eliminate the legal expectation of\n confidentiality. \n \n This is a global issue.\n \n The UK is an example, as recent cases demonstrate. In the case of\n UK v Secretary of State for the Home Department [2026] UKUT\n 81 (Hamid), the Upper Tribunal (Immigration and Asylum\n Chamber) delivered the first decision by an English court or\n tribunal to directly address what the legal professional\n privilege risks when confidential and privileged material is\n uploaded to open-source AI tools. \n \n It is important to remember that jurisdictions such as\n Singapore share the common law traditions of the UK and US, for\n example. As case law builds up around the world, the AI\n notetaking issue will become a global one. \n \n One of the points coming out of all this is that – as Severson\n said in the webinar – AI is a valuable tool. The ability to\n summarize and collect data is useful, to give just one\n example. The arrival of AI is, it should be said, also a\n reason why the grunt work that junior lawyers used to do is being\n replaced. These recent cases are also a reminder that large\n AI models, which rely on vast amounts of data, are, in a\n way, a sort of “public” field. \n \n Fintechs are starting to address the problem of how public AI can\n be. In\n early June, Custodia, a Swiss privacy-first AI startup, said\n it had launched Sentinel, a “physical AI thinking appliance\n developed and designed from the ground up for executives, family\n offices, scientific researchers, and any professional whose\n intellectual property is too valuable and too sensitive to trust\n to the cloud.”\n \n Perhaps it is not a coincidence that Custodia is Swiss – the land\n of bank secrecy (albeit no longer on cross-border matters),\n where privacy is still highly prized, as it should\n be. \n \n In its press release about the launch, Custodia said: “Rather\n than relying on pre-trained knowledge, Sentinel ingests your\n documents, understands their context, and retrieves precisely the\n right information to ground every answer it gives. Load thousands\n of files – financial records, research papers, legal\n documents, corporate history, correspondence – and Sentinel\n draws only from that store of knowledge. No hallucination from\n unrelated internet data.”\n \n Well, that is the sales pitch, and this news service is looking\n into this area about private AI to see just how thick the walls\n of privacy really are. What is clear, however, is that if\n you are a lawyer or a client, or indeed a professional wealth\n manager, investor or professional figure having a confidential\n discussion, the AI recording gizmo should be turned off.\n Or, at the very least, the use of these devices must be\n clearly disclosed ahead of time, giving affected parties a chance\n to refuse.","excerpt":"Lawyers can put confidentiality and privilege at risk if AI notetakers are used in meetings – one of the risks of this burgeoning technology that HNW individuals need to understand. In thinking about all the use cases and risks that attend AI in\n today’s private client advisory sector, one term that ought to\n stick in the mind right now is “confidentiality.”\n \n For those who don’t follow the twists and turns of AI closely, as\n well as legal cases, it might not appear obvious that one of the\n dangers of AI if not used properly is the destruction of\n lawyer-client privilege.","extraction":{"state":"ok","confidence":0.9,"error":null,"explanation":"High confidence: full text extraction produced 5514 characters.","diagnostics_url":"/api/diagnose?url=https%3A//www.wealthbriefing.com/html/article.php/opinion-of-the-week%253A-beware-of-ai-notetaking-threat-to-confidentiality-","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 5514 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":5514,"summary_length":586,"usable_text_length":5514,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":5514,"summary_length":586}}},"display_formats":["compact","card","full","digest_section","json"]},"daily_stack_record":{"title":"OPINION OF THE WEEK: Beware Of AI Notetaking Threat To Confidentiality - Wealth Briefing","url":"https://www.wealthbriefing.com/html/article.php/opinion-of-the-week%3A-beware-of-ai-notetaking-threat-to-confidentiality-","summary":"Lawyers can put confidentiality and privilege at risk if AI notetakers are used in meetings – one of the risks of this burgeoning technology that HNW individuals need to understand. In thinking about all the use cases and risks that attend AI in\n today’s private client advisory sector, one term that ought to\n stick in the mind right now is “confidentiality.”\n \n For those who don’t follow the twists and turns of AI closely, as\n well as legal cases, it might not appear obvious that one of the\n dangers of AI if not used properly is the destruction of\n lawyer-client privilege.","source":"Wealth Briefing","date":"2026-07-26T07:07:10+00:00","content":"Lawyers can put confidentiality and privilege at risk if AI notetakers are used in meetings – one of the risks of this burgeoning technology that HNW individuals need to understand.\n In thinking about all the use cases and risks that attend AI in\n today’s private client advisory sector, one term that ought to\n stick in the mind right now is “confidentiality.”\n \n For those who don’t follow the twists and turns of AI closely, as\n well as legal cases, it might not appear obvious that one of the\n dangers of AI if not used properly is the destruction of\n lawyer-client privilege. In an industry where HNW and UHNW\n individuals and families understandably value their privacy, this\n is a big problem. \n \n Many of us who go on Zoom, Teams and other platforms for business\n meetings might be told in advance – as we should – that “this\n meeting is being recorded” – usually in a cheerful voice. For\n journalists who haven’t mastered the grinding skill of shorthand\n – as I did decades ago – all these recording tools are a\n godsend. \n \n But there is a downside.\n \n As noted by the law firm White & Case in an April 23 report,\n entitled Attorney-client privilege and work product in the\n age of generative AI, two cases illustrate the risks –\n with very different results. \n \n On 10 February this year, the US District Court for the Eastern\n District of Michigan in Warner v. Gilbarco,\n Inc denied a motion to compel production of documents\n that a self-represented litigant had prepared using a public AI\n chatbot. The court found that work product protection\n applied. \n \n A week later, on 17 February, the US District Court for the\n Southern District of New York in United States vs.\n Heppner reached the opposite result, ordering a criminal\n defendant to produce documents he had generated using another\n public AI chatbot while seeking legal advice. \n \n The Heppner case was also mentioned at a recent media webinar,\n which I attended. The event was hosted by law firm ArentFox Schiff.\n Sarah Severson, a partner, explained the risks of AI in meetings\n where there are not clear understandings about what is\n involved.\n \n Notes taken via AI could be “discoverable,” she continued.\n Automated transcripts and summaries are classed as digital\n documents that opposing counsel can subpoena in civil or criminal\n cases. \n \n Severson said lawyers should always disclose their use of AI to\n clients in their engagement letters. \n \n “AI does not owe clients a duty of confidentiality or\n accountability,” she said. \n \n The root of the problem is that cloud-based processing,\n third-party vendor data access, and terms of service permitting\n model training eliminate the legal expectation of\n confidentiality. \n \n This is a global issue.\n \n The UK is an example, as recent cases demonstrate. In the case of\n UK v Secretary of State for the Home Department [2026] UKUT\n 81 (Hamid), the Upper Tribunal (Immigration and Asylum\n Chamber) delivered the first decision by an English court or\n tribunal to directly address what the legal professional\n privilege risks when confidential and privileged material is\n uploaded to open-source AI tools. \n \n It is important to remember that jurisdictions such as\n Singapore share the common law traditions of the UK and US, for\n example. As case law builds up around the world, the AI\n notetaking issue will become a global one. \n \n One of the points coming out of all this is that – as Severson\n said in the webinar – AI is a valuable tool. The ability to\n summarize and collect data is useful, to give just one\n example. The arrival of AI is, it should be said, also a\n reason why the grunt work that junior lawyers used to do is being\n replaced. These recent cases are also a reminder that large\n AI models, which rely on vast amounts of data, are, in a\n way, a sort of “public” field. \n \n Fintechs are starting to address the problem of how public AI can\n be. In\n early June, Custodia, a Swiss privacy-first AI startup, said\n it had launched Sentinel, a “physical AI thinking appliance\n developed and designed from the ground up for executives, family\n offices, scientific researchers, and any professional whose\n intellectual property is too valuable and too sensitive to trust\n to the cloud.”\n \n Perhaps it is not a coincidence that Custodia is Swiss – the land\n of bank secrecy (albeit no longer on cross-border matters),\n where privacy is still highly prized, as it should\n be. \n \n In its press release about the launch, Custodia said: “Rather\n than relying on pre-trained knowledge, Sentinel ingests your\n documents, understands their context, and retrieves precisely the\n right information to ground every answer it gives. Load thousands\n of files – financial records, research papers, legal\n documents, corporate history, correspondence – and Sentinel\n draws only from that store of knowledge. No hallucination from\n unrelated internet data.”\n \n Well, that is the sales pitch, and this news service is looking\n into this area about private AI to see just how thick the walls\n of privacy really are. What is clear, however, is that if\n you are a lawyer or a client, or indeed a professional wealth\n manager, investor or professional figure having a confidential\n discussion, the AI recording gizmo should be turned off.\n Or, at the very least, the use of these devices must be\n clearly disclosed ahead of time, giving affected parties a chance\n to refuse.","confidence":0.9,"diagnostics_url":"/api/diagnose?url=https%3A//www.wealthbriefing.com/html/article.php/opinion-of-the-week%253A-beware-of-ai-notetaking-threat-to-confidentiality-","quality_bucket":"high","failure_kind":"none","retryable":false,"quality_reason":"High confidence: full text extraction produced 5514 characters.","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 5514 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":5514,"summary_length":586,"usable_text_length":5514,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":5514,"summary_length":586}},"tags":[]},"fallback_formats":["markdown","json","html"],"actions":{"read":"/item/46315","export_markdown":"/api/items/46315/export?format=markdown","export_json":"/api/items/46315/export?format=json","diagnose":"/api/diagnose?url=https%3A//www.wealthbriefing.com/html/article.php/opinion-of-the-week%253A-beware-of-ai-notetaking-threat-to-confidentiality-"},"formats":{"full":{"id":46315,"title":"OPINION OF THE WEEK: Beware Of AI Notetaking Threat To Confidentiality - Wealth Briefing","url":"https://www.wealthbriefing.com/html/article.php/opinion-of-the-week%3A-beware-of-ai-notetaking-threat-to-confidentiality-","source":"Wealth Briefing","author":null,"published_at":"2026-07-26T07:07:10+00:00","locale":"en","topic":"ai","tags":[],"excerpt":"Lawyers can put confidentiality and privilege at risk if AI notetakers are used in meetings – one of the risks of this burgeoning technology that HNW individuals need to understand. In thinking about all the use cases and risks that attend AI in\n today’s private client advisory sector, one term that ought to\n stick in the mind right now is “confidentiality.”\n \n For those who don’t follow the twists and turns of AI closely, as\n well as legal cases, it might not appear obvious that one of the\n dangers of AI if not used properly is the destruction of\n lawyer-client privilege.","full_text":"Lawyers can put confidentiality and privilege at risk if AI notetakers are used in meetings – one of the risks of this burgeoning technology that HNW individuals need to understand.\n In thinking about all the use cases and risks that attend AI in\n today’s private client advisory sector, one term that ought to\n stick in the mind right now is “confidentiality.”\n \n For those who don’t follow the twists and turns of AI closely, as\n well as legal cases, it might not appear obvious that one of the\n dangers of AI if not used properly is the destruction of\n lawyer-client privilege. In an industry where HNW and UHNW\n individuals and families understandably value their privacy, this\n is a big problem. \n \n Many of us who go on Zoom, Teams and other platforms for business\n meetings might be told in advance – as we should – that “this\n meeting is being recorded” – usually in a cheerful voice. For\n journalists who haven’t mastered the grinding skill of shorthand\n – as I did decades ago – all these recording tools are a\n godsend. \n \n But there is a downside.\n \n As noted by the law firm White & Case in an April 23 report,\n entitled Attorney-client privilege and work product in the\n age of generative AI, two cases illustrate the risks –\n with very different results. \n \n On 10 February this year, the US District Court for the Eastern\n District of Michigan in Warner v. Gilbarco,\n Inc denied a motion to compel production of documents\n that a self-represented litigant had prepared using a public AI\n chatbot. The court found that work product protection\n applied. \n \n A week later, on 17 February, the US District Court for the\n Southern District of New York in United States vs.\n Heppner reached the opposite result, ordering a criminal\n defendant to produce documents he had generated using another\n public AI chatbot while seeking legal advice. \n \n The Heppner case was also mentioned at a recent media webinar,\n which I attended. The event was hosted by law firm ArentFox Schiff.\n Sarah Severson, a partner, explained the risks of AI in meetings\n where there are not clear understandings about what is\n involved.\n \n Notes taken via AI could be “discoverable,” she continued.\n Automated transcripts and summaries are classed as digital\n documents that opposing counsel can subpoena in civil or criminal\n cases. \n \n Severson said lawyers should always disclose their use of AI to\n clients in their engagement letters. \n \n “AI does not owe clients a duty of confidentiality or\n accountability,” she said. \n \n The root of the problem is that cloud-based processing,\n third-party vendor data access, and terms of service permitting\n model training eliminate the legal expectation of\n confidentiality. \n \n This is a global issue.\n \n The UK is an example, as recent cases demonstrate. In the case of\n UK v Secretary of State for the Home Department [2026] UKUT\n 81 (Hamid), the Upper Tribunal (Immigration and Asylum\n Chamber) delivered the first decision by an English court or\n tribunal to directly address what the legal professional\n privilege risks when confidential and privileged material is\n uploaded to open-source AI tools. \n \n It is important to remember that jurisdictions such as\n Singapore share the common law traditions of the UK and US, for\n example. As case law builds up around the world, the AI\n notetaking issue will become a global one. \n \n One of the points coming out of all this is that – as Severson\n said in the webinar – AI is a valuable tool. The ability to\n summarize and collect data is useful, to give just one\n example. The arrival of AI is, it should be said, also a\n reason why the grunt work that junior lawyers used to do is being\n replaced. These recent cases are also a reminder that large\n AI models, which rely on vast amounts of data, are, in a\n way, a sort of “public” field. \n \n Fintechs are starting to address the problem of how public AI can\n be. In\n early June, Custodia, a Swiss privacy-first AI startup, said\n it had launched Sentinel, a “physical AI thinking appliance\n developed and designed from the ground up for executives, family\n offices, scientific researchers, and any professional whose\n intellectual property is too valuable and too sensitive to trust\n to the cloud.”\n \n Perhaps it is not a coincidence that Custodia is Swiss – the land\n of bank secrecy (albeit no longer on cross-border matters),\n where privacy is still highly prized, as it should\n be. \n \n In its press release about the launch, Custodia said: “Rather\n than relying on pre-trained knowledge, Sentinel ingests your\n documents, understands their context, and retrieves precisely the\n right information to ground every answer it gives. Load thousands\n of files – financial records, research papers, legal\n documents, corporate history, correspondence – and Sentinel\n draws only from that store of knowledge. No hallucination from\n unrelated internet data.”\n \n Well, that is the sales pitch, and this news service is looking\n into this area about private AI to see just how thick the walls\n of privacy really are. What is clear, however, is that if\n you are a lawyer or a client, or indeed a professional wealth\n manager, investor or professional figure having a confidential\n discussion, the AI recording gizmo should be turned off.\n Or, at the very least, the use of these devices must be\n clearly disclosed ahead of time, giving affected parties a chance\n to refuse.","reading_time_min":4,"extraction":{"state":"ok","confidence":0.9,"error":null,"explanation":"High confidence: full text extraction produced 5514 characters.","diagnostics_url":"/api/diagnose?url=https%3A//www.wealthbriefing.com/html/article.php/opinion-of-the-week%253A-beware-of-ai-notetaking-threat-to-confidentiality-","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 5514 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":5514,"summary_length":586,"usable_text_length":5514,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":5514,"summary_length":586}}},"quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 5514 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":5514,"summary_length":586,"usable_text_length":5514,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":5514,"summary_length":586}},"actions":{"read":"/item/46315","export_markdown":"/api/items/46315/export?format=markdown","export_json":"/api/items/46315/export?format=json","diagnose":"/api/diagnose?url=https%3A//www.wealthbriefing.com/html/article.php/opinion-of-the-week%253A-beware-of-ai-notetaking-threat-to-confidentiality-"}},"digest":{"id":46315,"title":"OPINION OF THE WEEK: Beware Of AI Notetaking Threat To Confidentiality - Wealth Briefing","url":"https://www.wealthbriefing.com/html/article.php/opinion-of-the-week%3A-beware-of-ai-notetaking-threat-to-confidentiality-","source":"Wealth Briefing","topic":"ai","published_at":"2026-07-26T07:07:10+00:00","excerpt":"Lawyers can put confidentiality and privilege at risk if AI notetakers are used in meetings – one of the risks of this burgeoning technology that HNW individuals need to understand. In thinking about all the use cases and risks that attend AI in today’s private client advisory…","quality_bucket":"high","quality_reason":"High confidence: full text extraction produced 5514 characters.","reading_time_min":4,"cluster_id":null},"card":{"display_title":"OPINION OF THE WEEK: Beware Of AI Notetaking Threat To Confidentiality - Wealth Briefing","subtitle":"Wealth Briefing · 2026-07-26","summary":"Lawyers can put confidentiality and privilege at risk if AI notetakers are used in meetings – one of the risks of this burgeoning technology that HNW individuals need to understand. In thinking about all the use cases…","badges":["quality:high"],"links":{"read":"/item/46315","original":"https://www.wealthbriefing.com/html/article.php/opinion-of-the-week%3A-beware-of-ai-notetaking-threat-to-confidentiality-","diagnose":"/api/diagnose?url=https%3A//www.wealthbriefing.com/html/article.php/opinion-of-the-week%253A-beware-of-ai-notetaking-threat-to-confidentiality-"},"quality_warning":null},"export":{"title":"OPINION OF THE WEEK: Beware Of AI Notetaking Threat To Confidentiality - Wealth Briefing","url":"https://www.wealthbriefing.com/html/article.php/opinion-of-the-week%3A-beware-of-ai-notetaking-threat-to-confidentiality-","summary":"Lawyers can put confidentiality and privilege at risk if AI notetakers are used in meetings – one of the risks of this burgeoning technology that HNW individuals need to understand. In thinking about all the use cases and risks that attend AI in\n today’s private client advisory sector, one term that ought to\n stick in the mind right now is “confidentiality.”\n \n For those who don’t follow the twists and turns of AI closely, as\n well as legal cases, it might not appear obvious that one of the\n dangers of AI if not used properly is the destruction of\n lawyer-client privilege.","source":"Wealth Briefing","date":"2026-07-26T07:07:10+00:00","content":"Lawyers can put confidentiality and privilege at risk if AI notetakers are used in meetings – one of the risks of this burgeoning technology that HNW individuals need to understand.\n In thinking about all the use cases and risks that attend AI in\n today’s private client advisory sector, one term that ought to\n stick in the mind right now is “confidentiality.”\n \n For those who don’t follow the twists and turns of AI closely, as\n well as legal cases, it might not appear obvious that one of the\n dangers of AI if not used properly is the destruction of\n lawyer-client privilege. In an industry where HNW and UHNW\n individuals and families understandably value their privacy, this\n is a big problem. \n \n Many of us who go on Zoom, Teams and other platforms for business\n meetings might be told in advance – as we should – that “this\n meeting is being recorded” – usually in a cheerful voice. For\n journalists who haven’t mastered the grinding skill of shorthand\n – as I did decades ago – all these recording tools are a\n godsend. \n \n But there is a downside.\n \n As noted by the law firm White & Case in an April 23 report,\n entitled Attorney-client privilege and work product in the\n age of generative AI, two cases illustrate the risks –\n with very different results. \n \n On 10 February this year, the US District Court for the Eastern\n District of Michigan in Warner v. Gilbarco,\n Inc denied a motion to compel production of documents\n that a self-represented litigant had prepared using a public AI\n chatbot. The court found that work product protection\n applied. \n \n A week later, on 17 February, the US District Court for the\n Southern District of New York in United States vs.\n Heppner reached the opposite result, ordering a criminal\n defendant to produce documents he had generated using another\n public AI chatbot while seeking legal advice. \n \n The Heppner case was also mentioned at a recent media webinar,\n which I attended. The event was hosted by law firm ArentFox Schiff.\n Sarah Severson, a partner, explained the risks of AI in meetings\n where there are not clear understandings about what is\n involved.\n \n Notes taken via AI could be “discoverable,” she continued.\n Automated transcripts and summaries are classed as digital\n documents that opposing counsel can subpoena in civil or criminal\n cases. \n \n Severson said lawyers should always disclose their use of AI to\n clients in their engagement letters. \n \n “AI does not owe clients a duty of confidentiality or\n accountability,” she said. \n \n The root of the problem is that cloud-based processing,\n third-party vendor data access, and terms of service permitting\n model training eliminate the legal expectation of\n confidentiality. \n \n This is a global issue.\n \n The UK is an example, as recent cases demonstrate. In the case of\n UK v Secretary of State for the Home Department [2026] UKUT\n 81 (Hamid), the Upper Tribunal (Immigration and Asylum\n Chamber) delivered the first decision by an English court or\n tribunal to directly address what the legal professional\n privilege risks when confidential and privileged material is\n uploaded to open-source AI tools. \n \n It is important to remember that jurisdictions such as\n Singapore share the common law traditions of the UK and US, for\n example. As case law builds up around the world, the AI\n notetaking issue will become a global one. \n \n One of the points coming out of all this is that – as Severson\n said in the webinar – AI is a valuable tool. The ability to\n summarize and collect data is useful, to give just one\n example. The arrival of AI is, it should be said, also a\n reason why the grunt work that junior lawyers used to do is being\n replaced. These recent cases are also a reminder that large\n AI models, which rely on vast amounts of data, are, in a\n way, a sort of “public” field. \n \n Fintechs are starting to address the problem of how public AI can\n be. In\n early June, Custodia, a Swiss privacy-first AI startup, said\n it had launched Sentinel, a “physical AI thinking appliance\n developed and designed from the ground up for executives, family\n offices, scientific researchers, and any professional whose\n intellectual property is too valuable and too sensitive to trust\n to the cloud.”\n \n Perhaps it is not a coincidence that Custodia is Swiss – the land\n of bank secrecy (albeit no longer on cross-border matters),\n where privacy is still highly prized, as it should\n be. \n \n In its press release about the launch, Custodia said: “Rather\n than relying on pre-trained knowledge, Sentinel ingests your\n documents, understands their context, and retrieves precisely the\n right information to ground every answer it gives. Load thousands\n of files – financial records, research papers, legal\n documents, corporate history, correspondence – and Sentinel\n draws only from that store of knowledge. No hallucination from\n unrelated internet data.”\n \n Well, that is the sales pitch, and this news service is looking\n into this area about private AI to see just how thick the walls\n of privacy really are. What is clear, however, is that if\n you are a lawyer or a client, or indeed a professional wealth\n manager, investor or professional figure having a confidential\n discussion, the AI recording gizmo should be turned off.\n Or, at the very least, the use of these devices must be\n clearly disclosed ahead of time, giving affected parties a chance\n to refuse.","confidence":0.9,"diagnostics_url":"/api/diagnose?url=https%3A//www.wealthbriefing.com/html/article.php/opinion-of-the-week%253A-beware-of-ai-notetaking-threat-to-confidentiality-","quality_bucket":"high","failure_kind":"none","retryable":false,"quality_reason":"High confidence: full text extraction produced 5514 characters.","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 5514 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":5514,"summary_length":586,"usable_text_length":5514,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":5514,"summary_length":586}},"tags":[],"format_contract_version":"news_item_formats.v1"}}}