U.S. – How the US plans to protect older people from AI chatbots: the Federal Senior Chatbot Protection Act

US Senators Mark Kelly and Jim Justice have introduced a bill to protect older people from the risks posed by AI chatbots. It imposes a number of rules about safety, transparency and deception on the entities that make these AI systems available.

The Federal Senior Chatbot Protection Act aims to protect older Americans from the risk posed by AI chatbots. To mitigate these risks, the bill imposes five key obligations, such as requiring chatbots to always make it clear that they are AI. In addition, there is a further annual reporting obligation. Finally, enforcement will be carried out by the Federal Trade Commission (FTC) and the state attorneys general, with fines possible.

In July 2026, US Senators Mark Kelly and Jim Justice introduced the Federal Senior Chatbot Protection Act. This bill aims to establish a federal standard for the protection of older Americans who use AI chatbots. The bill is a response to a rapidly growing reality: millions of older people use chatbots for companionship, health advice, financial enquiries and day-to-day support. However, this same technology is increasingly being misused for manipulation, fraud and deception. This bill aims to reduce this risk by enshrining transparency, privacy protection, crisis response and a ban on manipulative design in law. 

Although the bill has attracted bipartisan interest, its legislative prospects remain uncertain. As of August 2026, the Federal Senior Chatbot Protection Act is still at the initial stage of the congressional process, having only been introduced and referred to the Senate Committee on Commerce, Science and Transportation. Legislative trackers currently estimate a low likelihood of enactment, reflecting both the early status of the bill and the broader challenges facing AI-related regulation in the United States. Industry resistance and competing congressional priorities further complicate its path forward, making it unclear whether the proposal will advance beyond committee consideration. 

1. Why this law is being introduced

Older adults are increasingly using AI chatbots for crucial tasks: managing medication, scheduling doctor’s appointments, keeping track of financial transactions, seeking social contact and even making legal decisions such as estate planning. When a chatbot sounds human, responds empathetically or presents itself as an expert, there is a risk of over-reliance, because they will place too much trust in a system that is not human, has no professional qualifications and bears no responsibility. This bill aims to establish consumer transparency and protection requirements for AI chatbots.

2. What the bill does 

The bill imposes 5 key obligations on any person who makes an AI chatbot available:

1. Mandatory transparency: It must always be clear that AI is involved, this means that the bill requires chatbots to state clearly and repeatedly that they are AI and not human. Many chatbots are designed to sound human, which can lead to confusion amongst older people. Additionally, the chatbots must state that they do not hold a professional licence. Chatbots must not pose as a doctor, therapist, lawyer, financial adviser or any other recognised professional.
2. Additional warnings for high-stakes decisions: When a chatbot detects that a user is seeking advice on a high-risk topic, the system must immediately warn the user that it does not provide a professional service. The bill defines high-stakes decisions as including, amongst others: medical decisions, estate planning, financial transactions and guardianship matters. The chatbot must actively encourage the user to consult a human professional. 
3. Crisis detection and mandatory referral: Chatbots must be capable of detecting crises: signs of suicidal ideation, medical emergencies, intentions to self-harm or a danger to others. Upon detection, the chatbot must not provide advice on medication, medical procedures or self-help. The system must immediately refer the user to the emergency services, a human professional or a crisis service provider, including the 988 Suicide and Crisis Lifeline. 
4. Data protection: The bill imposes strict restrictions on the collection, processing, storage and sharing of data obtained from a conversation with the user. Companies may only use data where strictly necessary for the provision of the service, legal obligations, preparation of reports or protection against fraud. Moreover, no AI training is allowed without explicit consent and users must be able to easily delete their conversation history.
5. Deceptive or manipulative interaction design practices: The bill prohibits design practices that, for example, encourage excessive reliance on such a chatbot, discourage users from seeking assistance or support from family members, caregivers, licensed professionals or other real-world relationships or encourage users to prioritise interactions with the chatbot over real-world relationships or activities. 

An entity that makes an AI chatbot available must prepare an annual report on all incidents involving the chatbot in high-stakes decisions or crisis situations. The report must specify the total number of material adverse incidents, describe the crisis detection methods used by the entity and outline any corrective actions taken. The Federal Trade Commission shall prescribe the standards for the format and submission of the report.

The Senior Chatbot Protection Act grants enforcement powers to the Federal Trade Commission (FTC) and to state attorneys general. The Federal Trade Commission exercises its powers as provided under the Federal Trade Commission Act, including the ability to impose civil fines, conduct investigations and compel compliance. Before initiating an enforcement action, the FTC must provide the entity with a 60-day cure period to remedy the violation.

State attorneys general may pursue enforcement actions when older residents within their jurisdiction have been misled or harmed. In addition, the Act authorises civil penalties of up to $50.000 per violation. 

The Federal Senior Chatbot Protection Act is an innovative legislative proposal. It recognises that older people are facing specific risks and that AI chatbots can pose a new source of deception, manipulation and even danger. By mandating transparency, crisis detention, data protection and anti-manipulation safeguards, the bill sets out a potential new standard for responsible AI design. Yet it remains important to underline that this is still a bill and not enacted law. Its introduction marks a significant first step, but its future depends on the legislative process and political momentum. Rather tan a final regulatory solution, the proposal opens a broader discussion on how societies should regulate AI when technology impacts human vulnerability.

This article was written by Hanne Wallaert as part of her internship at the The KU Leuven Centre for IT & IP Law. 

About Hanne Wallaert
Hanne Wallaert earned her Bachelor’s degree in Law from KU Leuven. She wrote her bachelor’s thesis on the extent to which the GDPR can prevent hacking. In September 2026, she will begin her Master’s degree in Law, where she will focus on economic and tax law. Within the field of IT law, she is particularly interested in data protection law.