AI at Sasphire
Our Use of Artificial Intelligence (AI), Including Generative AI, at Sasphire Legal
At Sasphire Legal, we are committed to integrating best-in-class technology into our legal services. This includes the strategic adoption of artificial intelligence (AI) and generative AI to enhance the quality, efficiency and consistency of the services we provide to our clients.
This page details our approach to AI, the tools we use, answers to frequently asked questions, and our governance contacts should you have any questions about how we select and use AI technologies.
How We Use AI
Sasphire Legal leverages AI in various ways, including to drive efficiencies in how we work and to improve the accuracy, consistency and quality of our work product. We currently use AI-powered tools to assist us in the conduct of legal research, document review and due diligence, data analysis, automation, and drafting. These are in addition to general business technologies that may incorporate AI.
Sasphire Legal has onboarded and approved the use of generative AI tools for the conduct of legal work. Currently approved tools include Harvey, a generative AI platform purpose-built for legal professionals. Harvey is used to assist with tasks including legal research, document review, drafting, summarisation and analysis. Our approved toolkit may evolve over time as new tools are evaluated and onboarded in accordance with our governance framework.
Our Approach to AI Governance
Sasphire Legal closely tracks the developing generative AI market and any regulations that apply to it. We maintain a rigorous approach to the adoption and use of solutions leveraging generative AI, and only do so after we have satisfied ourselves that we will remain in compliance with all regulatory requirements, relevant data security and controls requirements, and client terms and commitments.
Our governance framework includes the following commitments:
- We maintain internal policies, procedures and controls for the use of generative AI technologies within the firm.
- Our internal policies prohibit the use of confidential client data with public generative AI tools and govern the appropriate use of approved generative AI tools, including how our supervision obligations apply to AI-supported work.
- We only consider and permit the use of selected and approved generative AI tools that:
- appropriately secure our and our clients’ data;
- have been onboarded in accordance with our procurement and risk assessment processes;
- do not use our or our clients’ data to train underlying foundational models;
- encrypt our and our clients’ data both in transit and at rest; and
- process our and our clients’ data in compliance with all applicable data privacy laws.
- We implement platform-specific terms of use and provide training and guidance materials that support responsible and compliant usage of approved generative AI applications.
- All partners and staff are required to complete mandatory training covering how to use AI in compliance with our policies and our legal and professional obligations.
Data Security
Any client data that is used in connection with Sasphire Legal’s approved generative AI tools is stored only on secure, segregated servers. Client data is not co-mingled with data from any other person, and our clients’ data is not used to train underlying shared AI models. Data is encrypted both in transit and at rest.
As a matter of policy, Sasphire Legal does not approve the use of AI tools where our or our clients’ data is co-mingled with data from any other person or if our or our clients’ data is used to train underlying shared AI models.
We are particularly conscious of the sensitivity of employment-related personal information that may be processed through AI tools, including information relating to workplace disputes, health conditions, performance matters, and disciplinary proceedings. Where appropriate, additional safeguards such as de-identification protocols or restricted-access controls may be applied to such information.
Professional Obligations and Human Oversight
Our professional obligations are the same whether client work is produced with or without the use of any AI-based assistance. The use of approved AI tools does not in any way modify or reduce the firm’s responsibility for the quality, accuracy and completeness of advice and services provided by the firm, nor diminish our duties and professional obligations to our clients (in particular, but not limited to, the firm’s duties and obligations with respect to confidentiality).
AI is used as an assistive tool only. It does not make legal decisions, exercise legal judgment, or provide legal advice independently. All AI-assisted work product is reviewed and approved by a qualified legal practitioner before being provided to clients. All legal advice remains the professional judgment of our lawyers. This human oversight is particularly important in our employment and workplace relations practice, where matters frequently involve nuance, contextual judgment, and sensitive personal circumstances that require careful professional consideration.
Legal Professional Privilege
Privilege will not be lost solely because privileged information is passed through an AI tool. We consider this topic in relation to all AI tools evaluated and onboarded, and work with our providers to ensure data inputs and outputs are not used to train underlying shared models and are not shared with or disclosed to third parties or retained by the provider. Use of AI on each matter is carefully considered to ensure activities are structured appropriately so that privilege is maintained where relevant.
Frequently Asked Questions
What does Sasphire Legal use AI for?
Sasphire Legal leverages AI in various ways, including to drive efficiencies in how we work and to improve the accuracy, consistency and quality of our work product. We use AI-powered tools to assist in the conduct of legal research, document review and due diligence, data analysis, automation, and drafting.
Does Sasphire Legal use generative AI or large language models (LLMs)?
Yes. Currently approved generative AI tools include Harvey, a generative AI platform built for legal professionals. Harvey is used to assist with tasks including research, document review, drafting, summarisation, translation and analysis. Our approved toolkit may evolve over time as new tools are evaluated and onboarded in accordance with our governance framework.
Does Sasphire Legal use ChatGPT or other free generative AI tools on client matters?
Partners and staff may not input any Sasphire Legal or client personal or confidential data into free or publicly available generative AI tools. Failure to comply may result in disciplinary action.
Does Sasphire Legal use AI tools that use client data to train their models?
No. As a matter of policy, Sasphire Legal does not approve the use of AI tools where our or our clients’ data is co-mingled with data from any other person or if our or our clients’ data is used to train underlying shared AI models.
Where is client data stored when AI tools are used?
Any client data that is used in connection with Sasphire Legal’s approved generative AI tools is stored only on secure, segregated servers. Where relevant, we maintain storage on geo-located servers to meet the data residency requirements of our clients.
How does Sasphire Legal manage the risks associated with using AI tools?
Sasphire Legal maintains a risk framework relating to its adoption and use of AI, and takes such mitigating steps as are deemed appropriate to help manage identified risks. Use of all technology-based tools is subject to our policies on data privacy, information security and supervision, as well as tool-specific user terms and conditions. We also require all partners and staff to complete mandatory training covering how to use AI in compliance with these policies and our legal and professional obligations.
How does the use of AI affect the privileged status of client advice or work product?
Privilege will not be lost solely because privileged information is passed through an AI tool. We work with our providers to ensure data inputs and outputs are not used to train underlying shared models and are not shared with or disclosed to third parties or retained by the provider. Use of AI on each matter is carefully considered to ensure activities are structured appropriately to maintain privilege where relevant.
How is Sasphire Legal complying with its professional obligations when supervising work conducted with AI assistance?
Our obligations are the same whether client work is produced with or without the use of any AI-based assistance. The use of approved AI tools does not in any way modify or reduce the firm’s responsibility for the quality, accuracy and completeness of advice and services provided by the firm, nor diminish our duties and professional obligations to our clients.
Notification, Approval and Client Choice
Where relevant, we will comply with our clients’ notification and approval requirements regarding the use of any new AI-based tool prior to their use with those clients.
We recognise that some clients, particularly those with matters involving highly sensitive personal information (such as sexual harassment investigations, discrimination complaints, whistleblower matters, or sensitive workplace investigations), may prefer that AI tools are not used on their matter. Clients may request that AI not be used in connection with their matter at any time, and we will respect and comply with such instructions. If you have any questions about how AI may be used on your matter, or if you wish to discuss limitations on its use, please raise this with your usual Sasphire Legal contact.
Contact Us
For more information or support with understanding Sasphire Legal’s approach to and use of AI, please contact admin@sasphire.com.au.
For any other queries, please reach out to your usual Sasphire Legal contact.