‘Living Wishes Protocol’ Uses AI to Support Guardians in Making Education and Healthcare Decisions for Children, While Keeping Final Authority with Human Caregivers.
An Abu Dhabi-based mother is drafting a will that would instruct her son’s future guardians to seek guidance from ChatGPT when making important decisions about his upbringing if she passes away while he is still a minor. The unusual arrangement could raise new questions about the potential role of artificial intelligence in estate planning and child guardianship.
The proposed ‘Living Wishes Protocol’ would enable appointed guardians to consult a designated artificial intelligence system, currently named as ChatGPT in the draft document, together with a written account of the mother’s personal values, beliefs, parenting principles and aspirations for her son.
Under the proposed arrangement, the AI would neither assume the role of a legal guardian nor possess any decision-making authority over the child. Instead, it would function as a supporting advisory resource, helping human guardians interpret and apply the mother’s documented wishes when facing circumstances she may not have foreseen.
The will is currently being prepared for submission and registration with the Abu Dhabi Judicial Department. However, the document has not yet been officially registered or received court approval.
Byron James of Expatriate Law, the firm responsible for preparing the document, told Khaleej Times that the initiative represents an unconventional approach to incorporating artificial intelligence into guardianship arrangements and preserving parental wishes.
He explained that the intention is not to digitally recreate the mother or allow AI to take responsibility for raising her child. Rather, the objective is to retain elements of her parental guidance while ensuring that all important decisions remain under human control.
How Would the Living Wishes Protocol Work?
The seven-page bilingual draft, examined by, outlines the Living Wishes Protocol as a system designed to transform a parent’s recorded preferences into a flexible framework that can support guardians over time, instead of relying solely on a traditional written letter.
Under the proposed arrangement, appointed guardians would be expected to consult the designated AI platform before making major decisions affecting the child’s life.
These decisions could include selecting or changing schools, determining educational pathways, choosing where the child should live, considering non-emergency medical or psychological treatment, making decisions about religious upbringing and maintaining relationships with the mother’s relatives.
To receive guidance, guardians would submit the mother’s Living Wishes Record, a document detailing her personal values, beliefs, parenting philosophy and aspirations for her son, along with relevant information about the particular circumstances.
The AI system would then serve as an advisory resource, helping guardians consider how the mother’s documented principles might relate to the decision being made.
However, the guardian would retain complete responsibility and final authority over every decision concerning the child.
The draft clearly establishes that the AI would not be legally appointed under the will and would not receive any guardianship powers or authority over the child’s welfare.
It also specifies that the technology must never be portrayed as representing the mother’s voice or possessing knowledge of the decisions she would personally have made.
James explained that the central purpose of the proposal is to provide future caregivers with a more useful source of parental guidance if the mother dies while her son is still young.
Although parents can document their beliefs, priorities and hopes for their children, they cannot predict every situation that may arise throughout the following 10 to 15 years.
The proposed framework therefore seeks to help guardians interpret those recorded wishes in changing circumstances, rather than treating them as fixed instructions.
It also aims to ensure that the mother’s preferences do not become inflexible rules that prevent the child’s own needs, circumstances and developing independence from being considered as he grows older.
Under the proposed framework, guardians would be required to consider the mother’s documented principles alongside the child’s preferences, emotions and individual needs, while also exercising their own judgement about his overall wellbeing.
As the child grows older and develops a greater ability to communicate his views, those opinions would carry increasing importance in decisions affecting his life.
The arrangement also includes an annual review process, continuing until the child reaches the age of 18. During these assessments, guardians would evaluate his academic progress, physical and emotional health, happiness, friendships, personal interests and evolving needs.
The reviews would also examine whether the mother’s Living Wishes Record continues to reflect the child’s circumstances and remains suitable as a source of guidance.
Once the child reaches an appropriate level of maturity, he would be encouraged to participate in these discussions and express his own preferences.
Additionally, the draft proposes appointing an independent human adviser to assess the protocol at least once every year or whenever requested by the guardian.
The adviser’s responsibilities would include evaluating whether the arrangement adequately protects the child’s privacy, respects his growing independence and continues to use an appropriate AI system.
AI Protocol Would Not Apply During Emergencies
The proposed Living Wishes Protocol would not be used in situations requiring immediate action to protect the child’s life, health or personal safety.
The framework would also be prohibited from supporting decisions that conflict with existing laws or court orders.
Importantly, the draft clarifies that a guardian’s failure to consult or follow the protocol would not automatically make a decision legally invalid or result in the cancellation of their guardianship appointment.
To maintain accountability, guardians would be expected to preserve confidential documentation of important AI consultations.
These records would include the questions submitted to the system, the responses generated and the decisions ultimately taken by the guardians.
Where guardians choose a course of action different from the AI’s recommendations, they would also be expected to document their reasoning.
Such records could be provided to the estate’s executors, the appointed independent adviser or a court when formally requested.
Concerns Over Children’s Privacy and Data Protection
The proposed use of artificial intelligence in guardianship decisions has also raised important questions about protecting children’s personal information.
James acknowledged that several technical aspects of the arrangement, including how the mother’s recorded wishes would be accessed and stored, have yet to be finalised.
He explained that the legal framework should not become unnecessarily dependent on a specific commercial AI account, software version or technology provider, particularly because the arrangement may need to remain functional for many years.
According to James, preserving the mother’s documented values and parenting guidance is more important than ensuring the continued use of any particular AI platform.
Further consideration is also needed regarding how personal information would be securely stored, who would be authorised to access it and what details might be shared with external AI service providers.
The legal team is also exploring ways to limit the disclosure of sensitive information, including removing identifying details or anonymising personal records where possible.
James emphasised that highly sensitive details about a child’s private life should not be unnecessarily submitted to commercial artificial intelligence services.
He maintained that protecting the child’s privacy, wellbeing and growing ability to make independent decisions must remain the priority, even when the purpose of the arrangement is to preserve parental guidance.
Why Is This Proposal Significant?
The proposed will stands out not merely because it references ChatGPT, but because it attempts to establish a continuing advisory process through which guardians could use artificial intelligence to interpret a deceased parent’s documented wishes.
Rather than treating parental instructions as a fixed document, the arrangement would allow guardians to revisit those values as the child’s circumstances change over time.
James said Expatriate Law was unaware of any previously registered will containing the same combination of features, including annual assessments, greater recognition of the child’s preferences as he matures, independent human supervision and explicit restrictions against AI presenting itself as the deceased parent.
However, he clarified that the firm was not claiming the arrangement to be the first of its kind worldwide.
Because wills and other testamentary documents are not publicly accessible in every jurisdiction, and no comprehensive international review has been conducted, such a claim cannot be conclusively established.
The proposal’s legal recognition in the UAE also remains uncertain.
James confirmed that the Living Wishes Protocol has not yet received judicial approval or formal acceptance, meaning its legal enforceability and treatment under UAE law have yet to be determined.
ChatGPT Has No Independent Legal Status
James explained that ChatGPT does not possess legal personhood, but emphasised that the proposed arrangement neither designates the AI system as a legal guardian nor grants it any authority to make decisions on behalf of the child.
However, he acknowledged that it remains uncertain whether the court will consider this distinction legally acceptable or approve the proposed provisions in their existing form.
The document may undergo further revisions following legal assessments or recommendations from the relevant authorities. These discussions could also determine which provisions should be incorporated into the officially registered will and which would be more appropriately included in a separate letter outlining the mother’s wishes.
The draft explicitly clarifies that it is intended solely for discussion and does not constitute a formally registered legal instrument. The sections relating to court registration and official attestation have also been left incomplete.
James added that the identities of the family members and the child’s age are being withheld to protect the minor’s privacy.
Despite its uncertain legal standing, the proposal introduces an important question for modern estate planning: How can parents ensure that their values and guidance continue to influence their children’s upbringing after their death, particularly when future circumstances and challenges cannot be anticipated?
The Abu Dhabi mother’s proposed approach seeks to address this challenge by combining human judgement with AI-assisted guidance rather than relying entirely on either one.
Under the proposed framework, artificial intelligence would serve only as a supporting resource, while legal authority would remain with human guardians, subject to judicial oversight and the child’s evolving needs, preferences and independence.
Ultimately, the arrangement aims to preserve a parent’s influence without allowing technology to replace human responsibility or override the child’s best interests.
