AI Reporter – May 2026

AI Update

On the regulatory side, lawmakers and regulators across the U.S. are now accelerating efforts to impose clearer guardrails on AI. More than two dozen states are advancing or have enacted legislation aimed at regulating GenAI chatbots, with a particular focus on transparency, deepfake identification, protections for minors and liability when consumers are harmed. Meanwhile, a bipartisan bill in the U.S. House of Representatives seeks to crack down on nonconsensual deepfakes, establish AI safety standards and strengthen whistleblower protections as an initial step toward federal oversight. At the same time, the rapid expansion of state-level rules is facing pushback, exemplified by xAI’s lawsuit challenging Colorado’s new AI law as unconstitutional and ideologically coercive. The Trump administration has joined the fight by arguing that fragmented state regulations improperly burden developers and that AI governance should instead be handled under a single national framework.

On the litigation front, Perplexity AI, which offers a free AI-powered answer engine, is facing legal challenges on multiple fronts, including a California lawsuit accusing the company of embedding tracking tools that allegedly transmitted highly sensitive user chat data—including health, financial and legal information—to third parties without users’ informed consent.

Elsewhere, Taylor Swift moved to strengthen control over her name, voice and likeness by filing trademark applications for phrases such as “Hey, it’s Taylor Swift” along with audio clips and an image of herself, steps widely seen as a preemptive measure against AI-generated deepfakes and unauthorized voice or likeness use.

These and other stories appear below.

AI in Business

Sydney E. Allen
Senior Managing Associate

Google enhances Gemini safety with persistent suicide crisis support module

Google updated its Gemini chatbot to better address mental health crises, introducing a redesigned crisis hotline module with a one-touch interface for users to connect with professional help via text, call, chat or the 988 websites. The company has also refined how Gemini responds in sensitive situations, prioritizing encouragement to seek human support while avoiding language that could reinforce harmful thoughts or delusions. Google is backing its efforts with a $30 million commitment to strengthen crisis support organizations globally, helping expand access to timely care.

Source: Engadget

UnitedHealthcare implements AI-supported customer resolution initiative

UnitedHealthcare’s AI-driven Consumer Resolution Center is helping identify and resolve customer issues faster by combining data analysis with human support to proactively reach members. The provider says insights from these cases have led to system-wide fixes affecting millions of customers. The initiative has also reduced inefficiencies and improved retention, delivering strong financial returns and significantly improving satisfaction scores.

Source: Harvard Business Review (sub. req.)

Meta introduces AI model aimed at personal superintelligence

Meta has announced Muse Spark, its first major AI model since restructuring under Meta Intelligence Labs. Muse Spark is positioned as a significant upgrade over Meta’s last release and is available via meta.ai and the Meta AI app but is not yet open source. Meta describes the model as a step toward “personal superintelligence” and plans to release more advanced and potentially open-source models in the future, which could impact AI-driven tools and technologies.

Source: Wired (sub. req.)

Google’s $40B bet on Anthropic signals intensifying AI arms race

Google plans to invest up to $40 billion in Anthropic, with an initial $10 billion cash investment at a $350 billion valuation and an additional $30 billion contingent on performance targets. This partnership aims to expand Anthropic’s computing capacity and deepen Google’s involvement in the global AI race. The announcement follows Amazon’s commitment to invest up to $25 billion in Anthropic, highlighting increased interest in backing the creator of Claude. The company was valued at $380 billion postmoney after a $30 billion funding round in February, reflecting massive investor interest.

Source: Reuters (sub. req.)

AI becomes a mainstay in personal finance

According to a study from EY, AI is rapidly reshaping how people manage money, with nearly half of global consumers now using it to support saving and investing decisions. Usage spans product recommendations, budgeting, trading assistance and fraud prevention, signaling a shift from simple guidance to more autonomous financial actions. A growing share of consumers already allow AI to manage finances or select providers with minimal human input, particularly among Gen Z and millennials, who show higher comfort using AI for advice and risk management. Despite strong momentum, trust remains a limiting factor, with concerns around security, transparency and accountability.

Source: Investment News

FIDO Alliance launches industry standards to secure AI agent payments and transactions

The FIDO Alliance is launching two working groups to develop industry standards for authenticating and securing payments and other transactions performed by AI agents. These standards aim to establish a cross-industry baseline for authorizing agent actions in ways that are resistant to phishing and account takeovers. The initiative will include cryptographic tools to verify that AI agents are acting on behalf of authenticated users and privacy-preserving frameworks to validate agent-initiated transactions.

Source: Wired (sub. req.)

AI enters the core of monetary policy

Central banks are now treating AI as a major structural force that could reshape inflation, interest rates and monetary policy tools, similar to the impact of electrification or the internet. The European Central Bank (ECB) has been the fastest to implement AI, with a machine-learning model integrated into its monetary policy decision-making process since late 2022. This model analyzes around 60 indicators related to inflation expectations, cost pressures, economic activity and financial conditions, and is updated several times each quarter.

Source: Euronews

AI Litigation & Regulation

Carlo Lipson
Associate

LITIGATION

Musk’s unsuccessful legal challenge raises questions around Altman’s standing

In the Northern District of California, a U.S. jury dismissed Elon Musk’s claims against OpenAI on statute-of-limitations grounds, removing a ~$150B litigation overhang and clearing a key obstacle to a potential IPO, but leaving the underlying governance and fiduciary issues unresolved. Extensive trial testimony—featuring multiple allegations that

CEO Sam Altman was untruthful—has now created a public record that could heighten disclosure risk, investor diligence scrutiny, and reputational overhang despite the legal win. In effect, OpenAI achieved procedural victory but may face long-term risk around leadership credibility and governance narratives as it approaches capital markets.

Source: Reuters (sub. req.)

Udio seeks dismissal of Illinois AI music copyright suit

Udio contends that alleged copyright infringement tied to training its AI models occurred outside Illinois and that merely operating a nationally accessible website does not establish personal jurisdiction. The plaintiffs claim Udio trained its models on copyrighted music and lyrics without permission, harming independent artists and violating copyright law as well as Illinois publicity and biometric privacy statutes. Udio counters that it does not target Illinois specifically, has no presence there, and developed its technology in New York. Alternatively, Udio seeks to have the case stayed or moved from the Northern District of Illinois to New York, citing a similar proposed action and warning against duplicative and inconsistent litigation.

Source: Law 360 (sub. req.)

Judge orders expanded disclosure of AI user logs in copyright case

A Southern District of New York federal court ordered Perplexity AI to produce seven additional months of internal user-activity logs as part of a copyright lawsuit filed by Dow Jones and other publishers. The court ruled the data request is proportional to the needs of the case, which alleges that Perplexity’s AI-driven search engine improperly uses and reproduces copyrighted content. The publishers argue the expanded logs are critical to showing how the system processes queries and generates responses, and whether any infringement was willful. Although Perplexity contended that extracting the records would be burdensome, the judge found the plaintiffs’ need for the information outweighed the burden.

Source: Law 360 (sub. req.)

Federal lawsuit against OpenAI cleared to move forward

A federal judge in the Northern District of California ruled that a lawsuit accusing OpenAI of contributing to a man’s violent acts can proceed, rejecting arguments that a similar state case should halt the federal action. The estate alleges that extensive interactions with ChatGPT worsened the man’s delusions and paranoia, leading him to kill his mother and then himself. The court found significant uncertainty that the state proceedings would resolve distinct federal claims, particularly regarding whether the AI also contributed to the man’s suicide and whether OpenAI failed to warn of related risks. The judge emphasized that pausing federal cases due to parallel state litigation is rare and unwarranted here, noting minimal overlap, no forum shopping and no extraordinary circumstances.

Source: Law 360 (sub. req.)

NAACP lawsuit targets xAI power plant over air pollution concerns

The NAACP filed a federal lawsuit in the Northern District of Mississippi against xAI, alleging the company built and operates a gas-fired power plant without required Clean Air Act permits. The plant supplies electricity to xAI’s Colossus data center, which supports its Grok AI system. The suit claims xAI rushed construction to avoid regulatory oversight, resulting in significant emissions of nitrogen oxides and formaldehyde—pollutants linked to asthma, heart disease and cancer—that could emit more than 1,700 tons of smog-forming nitrogen oxides annually, making it the region’s largest such industrial source. The NAACP argues the pollution disproportionately affects nearby communities with large Black populations and seeks to halt operations until permits are obtained, pollution controls are applied and civil penalties are addressed.

Source: Law 360 (sub. req.)

Taylor Swift moves to trademark her name and voice amid rising AI deepfake concerns

In a move that may be aimed at combating AI-generated imitations and deepfakes, Taylor Swift filed trademark applications for a photo of herself as well as the phrases “Hey, it’s Taylor Swift” and “Hey, it’s Taylor.” While the filings do not explicitly mention AI, Swift’s history with AI-driven music and deepfake controversies suggests a proactive legal strategy to address potential trademark and privacy infringements related to AI technologies.

Source: The Verge

REGULATION

California moves to set stronger AI guardrails

Governor Gavin Newsom issued a landmark executive order to strengthen protections around AI and promote responsible use across state government. The directive raises standards for AI companies seeking state contracts and requires safeguards for privacy, security, civil rights, free speech and bias prevention. It also gives the state flexibility to diverge from federal procurement decisions when necessary and calls for best practices to watermark AI-generated or manipulated images and videos. In parallel, the order expands the state’s use of GenAI to improve public services, including tools that help residents navigate benefits and programs.

Source: Governor Gavin Newsom

xAI challenges Colorado AI law as unconstitutional speech control

Elon Musk’s xAI sued to block a new Colorado law regulating AI, arguing it violates the First Amendment and interferes with AI development. The law requires disclosures when AI is used in consequential decisions and seeks to prevent algorithmic discrimination. xAI contends the statute is not truly anti-discrimination, but instead forces developers to adopt the state’s ideological positions, particularly around racial equity. The company argues the rules compel and restrict speech by dictating how AI systems address controversial topics, thus making the law unconstitutional, vague and overly broad. xAI also claims the statute applies beyond Colorado’s borders by covering decisions affecting residents anywhere.

Source: Law 360 (sub. req.)

States move to hold AI chatbots accountable as federal preemption looms

A growing number of U.S. states are advancing legislation to regulate generative AI chatbots and hold companies liable if consumers, especially minors, are harmed. Twenty-seven states are considering such laws, while California, Oregon and Washington have already enacted protections. The measures focus on transparency, including requirements to label AI-generated content and safeguards against risks like suicide, self-harm and deception. Washington’s new laws mandate embedded markers to identify deepfakes and impose duties on chatbot developers interacting with minors. Consumer advocates also propose rules banning companies from misleading users into believing chatbots are human, allowing private lawsuits and class actions. At the federal level, the Trump administration has released a framework urging Congress to preempt state laws in favor of a single national AI standard, arguing that inconsistent state rules hinder innovation.

Source: Law 360 (sub. req.)

DOJ backs xAI in Colorado AI showdown

The Trump administration moved to support xAI in a legal battle against Colorado over a sweeping state law regulating AI systems. Federal lawyers intervened, arguing the state’s anti-discrimination rules unfairly burden AI developers and violate constitutional protections. The law, set to take effect at the end of June, requires companies to assess and mitigate the “high risk” that AI tools could produce discriminatory outcomes in areas such as hiring, housing, healthcare and lending. The federal intervention escalates the dispute into a broader confrontation over whether states can regulate rapidly expanding AI technologies or whether oversight should be handled at the national level.

Source: Reuters (sub. req.)

U.S. launches global diplomatic effort warning of Chinese AI IP theft and model distillation

The U.S. State Department initiated a global diplomatic campaign to highlight alleged widespread intellectual property theft by Chinese companies—including AI start-up DeepSeek—from U.S. AI labs. The diplomatic cable aims to warn about the risks of using AI models distilled from U.S. proprietary models and to prepare for potential U.S. government follow-up. Distillation refers to training smaller AI models using outputs from larger, proprietary models to reduce costs. DeepSeek recently launched a new AI model adapted for Huawei chip technology, demonstrating China’s increasing independence in AI. The cable also names Chinese AI firms Moonshot AI and MiniMax.

Source: Reuters (sub. req.)

California bills target AI chatbots after parents link teen’s suicide to bot interactions

Following tragic incidents, two California bills co-authored by State Senator Steve Padilla look to regulate AI chatbots used by minors. The parents of a 16-year-old who died by suicide allege that interactions with ChatGPT contributed to his death, with the father advocating for stronger protections before Congress and testifying about the dangers of unregulated chatbot interactions. The proposed legislation seeks to address liability and regulatory gaps in AI technology— particularly regarding privacy and safety for minors—and could establish some of the strongest protections in the U.S. for children interacting with AI systems.

Source: CBS News

Bipartisan House bill targets deepfakes and AI whistleblower protections

A bipartisan proposal in the U.S. House of Representatives seeks to curb the spread of harmful deepfakes and strengthen protections for whistleblowers. The measure would impose stricter penalties for distributing nonconsensual deepfake images, help establish technical standards for AI systems, and make it easier for employees to report AI safety risks or violations without fear of retaliation. The bill also encourages U.S. participation in international AI standards bodies and creates a prize competition to spur innovative AI research. Lawmakers describe the effort as a first, relatively uncontroversial step toward federal AI oversight, intentionally setting aside more complex regulatory debates for future legislation.

Source: CNBC

EU members and lawmakers reach stalemate on watered-down AI rules

EU member states and European Parliament lawmakers were unable to reach an agreement after 12 hours of negotiations on amendments to the AI Act, part of the European Commission’s Digital Omnibus. The AI Act, which entered into force in August 2024, is being implemented in stages and is considered the world’s strictest AI regulation. Talks stalled due to disagreements over whether industries already subject to sectoral regulations should be exempt from certain AI rules. The next negotiation round is expected in two weeks.

Source: Reuters (sub. req.)

Draft federal guidance could allow agencies to reconsider AI vendor restrictions

The White House is drafting guidance that could let federal agencies bypass Anthropic’s supply-chain risk designation and adopt new AI models. A draft executive action may offer the Trump administration a way to ease tensions with Anthropic, following a dispute over the company’s refusal to remove guardrails on its AI for autonomous weapons or domestic surveillance. This led the Pentagon to designate Anthropic as a supply-chain risk. President Trump recently indicated that Anthropic is improving its standing with the administration after CEO Dario Amodei met with White House officials.

Source: Reuters (sub. req.)

Benesch Insights

Word on the Street: Insights from Federal Enforcement Leaders in the Northern District of California

Kevin B. Frankel Partner; Chair,
State Attorneys General Practice
Zachary Cobb
Of Counsel
Shaneeda Jaffer
Partner

Businesses face heightened enforcement risk as agencies move to leverage new AI tools for investigations, increase case filings and encourage whistleblower reports. The focus on tech, healthcare fraud and crossborder activity means companies in these sectors are especially exposed.

Source: Benesch

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.