We Protect Your Innovations

Last month the US Copyright Office released a pre-publication edition of the third part of their Copyright and Artificial Intelligence Report. The three part report covers the US Copyright Office's approach to digital replicas, copyrightability/authorship, and generative AI training. Click the links to read each part of their report.

On April 1, the USPTO announced a new guidance for SNQ determination in ex parte Reexamination Proceedings. Read their official notice on the matter here

The USPTO recently announced the release of the “Class ACT” AI assistant. The tool can immediately assign international classes to unclassified applications along with the design search codes and pseudo marks that make these records searchable. The USPTO is hoping that the tool will produce faster, more accurate results. Read the press release here.

Past News

USPTO launches Streamlined Claim Set Pilot Program , Form SB0472

USPTO launches AI-powered DesignVision search tool for design patent examiners

US Patent and Trademark Fees have increased in 2025. Follow the links below for more information.

US Patent Fees

• USPTO Patent Fee Schedule Increases 2025 and Other News
• Summary of Patent Fee Increases 2025

US Trademark Fees

• USPTO Trademark Fee Schedule Increases 2025 and Other News
• Summary of Trademark Fee Increases 2025

Last month there were two cases involving the use of copyrighted works to train large language models (LLMs). The judges in both Bartz v. Anthropic PBC and Kadrey v. Meta Platforms Inc. both determined that copying large copyrighted works to train the defendants' respective LLMs was considered tranformative use. However, the judges in both cases acknowledged potential avenues for infringement in future cases. In Bartz v. Anthropic, the judge deemed that the copyrighted works must be lawfully acquired in order for them to be considered as fair use, while in Kadrey v. Meta Platforms, the judge indicated that market dilution could create a roadblock for LLM trainers despite the plaintiff's lack of evidence to suggest such damages. Read the full court documents for Bartz v. Anthropic PBC here and Kadrey v. Meta Platforms Inc here.

Last month, Congress introduced H.R. 8094, the AI Foundation Model Transparency Act of 2026. The bill targets large AI models (Claude, ChatGPT), requiring their developers to disclose details on how their models are trained, what they are designed to do, potential risks in using them, safety standards, and computational resources. The bill does not serve to moderate or regulate the AI models, but merely help inform the public about their functions through transparency. Read the House Bill here.

The White House recently released a National Policy Framework for Artificial Intelligence. While the document does not enforce new regulations surrounding AI, its intent is to build a broad framework for nationwide guidelines. The framework, which can be read here, covers the following topics: 1) Protecting children and empowering parents, 2) Safeguarding and Strengthening American communities, 3) Respecting intellectual property rights and supporting creators, 4) preventing censorship/protecting free speech, 5) Enabling innovation and ensuring American AI dominance, 6) Educating Americans and developing an AI-ready workforce, and 7) Establishing a Federal policy framework to preempt cumbersome AI state laws.

On Jan 1, 2026, California’s Training Data Transparency Act took effect, establishing standards of transparency for GenAI tools. The bill aims to ensure that developers disclose information relating to the datasets used for training AI. The bill has already received a challenge in court from AI developer, xAI. Read California’s Assembly Bill 2013 here, and the court document for xAI’s complaint here.

Past AI News:

Attorneys General Form a Bipartisan AI Task Force – Read about the task force here.

Protecting Copyrighted Works Used for Training Generative AI Platforms; Read more on the bipartisan CLEAR Act here and the bill itself here.

Department of Labor releases AI Literacy Framework

Two recent court rulings show the fine line in how AI consulting affects attorney-client privilege. In the federal case of the United States v. Heppner, seized documents showing the defendant’s use of Anthropic’s Claude AI tool to prepare for the case was deemed to not be protected either attorney-client privilege or the work-product doctrine. This decision was based on the grounds that the defendant’s prompts and outputs using the AI tool were not directed by their legal counsel and the communication with a consumer-level AI tool was considered a third-party communication.

A civil case in Michigan, Warner v. Gilbarco Inc, took a different turn regarding AI generated documents and attorney-client privilege. In this case, the plaintiff used ChatGPT to prepare for litigation and the defendant sought production of any documents related to the plaintiff’s use of an AI tool, believing it to be a waiver of work product protection. The defendant was denied discovery of that information because the a waiver of work-product privilege would require disclosure to an adversary and due to ChatGPT being identified as a tool rather than a person.

Read the court documents United States v. Heppner Warner v. Gilbarco Inc.

Legislators in several states have begun proposing bills to help regulate AI chatbots to prevent them from being misused or providing misleading information to their users.

California’s SB 243 aims to require AI chatbots to clearly disclose that users are interacting with AI, while also creating safeguards when the AI agent knows that the user is a minor, and implementing reasonable measures to prevent the chatbot from providing harmful content. Failure to adhere to these protocols would amount to $1000 for each violation. Washington’s proposed bill, SB 5984, follows similar guidelines with more in-depth guidelines regarding the protection of minors. Oregon has just passed SB1546 which employs similar protocols introduced in the California and Washington bills, as well as providing resources to users when it detects any ideation of any self-harm. The Oregon bill would also allow victims of nefarious AI chatbots to pursue litigation to recover the great of either actual damages or $1000 for each violation.

New York legislators have also introduced bills that propose safeguards when interacting with minors. Bill S9051 prohibits sharing personal pronouns, expressing personal opinions, and simulating close relationships with the user. Additionally, State Senator Kristen Gonzalez sponsored bill S7263, which would bar AI chatbots from providing “any substantive response, information, or advice, or take any action” that “if taken by a neutral person” would be considered unauthorized practice of law, medicine, or other fields requiring licensure in the state.

Tennessee’s Senate has already passed SB1580, which prohibits AI chatbots from representing themselves as mental health professionals. State legislators also proposed SB1493, which includes harsher punishment for providing malicious content, making it a Class A felony for knowingly training AI to encourage self-harm or simulating emotional relationships with its users. The bill also prohibits chatbots from mimicking human behavior or encouraging the user to provide private and sensitive information such as finances and information regarding interpersonal relationships.

The widespread emergence of AI has prompted the Insurance Services Office (ISO) to introduce new endorsements for providers to adopt to help them navigate such risky new territory. By applying these endorsements to Commercial General Liability (CGL) policies, insurers can exclude coverage for any losses that can be linked to Generative AI. These exclusions consist of the following: 1) Exclusions relating to bodily injury, property damage, or personal and advertising injury, 2) Exclusions relating to personal and advertising injury only, and 3) Exclusions relating to products or completed operations. These exclusions should prompt policy holders to have a more proactive approach to not only understanding how AI is used in their operations, but also in recognizing how to navigate them to maximize their coverage when dealing with the inevitable risks that come with this constantly evolving technology.