Faegre Drinker Biddle & Reath LLP, a Delaware limited liability partnership | This website contains attorney advertising.
November 16, 2026

How the Midterm Results Could Reshape AI and Data Center Regulations, and More Developments

Artificial Intelligence Briefing

Table of Contents

  1. How the 2026 Midterms Could Reshape AI Regulation

  2. Rising State and Local Scrutiny Threatens Data Center Timelines

  3. Anthropic Prepares for Record IPO as Investors Push for Lower Valuation and Governance Concerns Mount

  4. Calling Plays from the Field: Anthropic Pitches AI Safety Bill

  5. FTC Orders Airlines, Gig Platforms, and Retailers to Explain Personalized Pricing Algorithms

  6. Awe and Anxiety as AI Races ahead in Mathematics

  7. Frontier AI Exposes Decade-Old Flaws at the Core of the Web

  8. Recent Binance Hack Highlights AI-Era Security Risks

  9. Banks Reportedly Find Unauthorized AI Agents in Customer Accounts and Move to Block Them

  10. FSB Finalizes Its Sound Practices for the Responsible Adoption of AI


This month's briefing covers how the new Congress could address AI regulation, as well as the rising state and local scrutiny of data centers. Meanwhile, Anthropic prepares for a record IPO. Read on for a deeper dive into these and more key updates.

Regulatory, Legislative & Litigation Developments

How the 2026 Midterms Could Reshape AI Regulation

Voter anxiety about artificial intelligence, among several factors, helped Democrats take control of both chambers of Congress: in an exit poll of 100,000 voters, 56% said AI and data centers were a "crucial or major factor" in their vote, and two-thirds of those voters backed Democrats. The incoming majority is promising hearings and "reasonable guardrails"; and committee chairs are expected to use their new subpoena and oversight powers on AI safety, model testing, and Big Tech's ties to the White House. Sweeping federal legislation still faces a likely presidential veto, so most analysts expect change to come first through oversight. Broad federal preemption of state AI laws also now looks unlikely, leaving businesses to manage a growing state-by-state patchwork.

Rising State and Local Scrutiny Threatens Data Center Timelines

Newly elected governors and local ballot measures are likely to drive increased state and local regulation, which could create longer timelines and greater uncertainty for data center development. Texas is a key area to watch, as the state was projected to supply 25% to 30% of US data center capacity and Governor-elect Gina Hinojosa has called for an immediate moratorium on new data center construction pending additional community protections. Similarly, governors-elect in Ohio, Georgia, Michigan, and New Mexico campaigned on pausing data center development, including a conditional pause backed by Ohio Governor-elect Amy Acton and a temporary pause backed by Georgia Governor-elect Keisha Lance Bottoms. Progressive proposals targeting data center power consumption could add another layer of scrutiny.

Anthropic Prepares for Record IPO as Investors Push for Lower Valuation and Governance Concerns Mount

Anthropic is preparing what could be the largest initial public offering in history, with prospective buyers now discussing a valuation of roughly $1.2 trillion to $1.3 trillion, well below earlier talk of $2 trillion, and a raise of as much as $64 billion. The company's registration statement includes more than 90 pages of risk disclosures, including warnings that its AI technology could pose catastrophic or existential risks to humanity. Through a single Class F share, Anthropic's seven co-founders will hold 50.1% of the vote, while public Class A shareholders receive one vote per share, capped at 49.9% combined, and the Long-Term Benefit Trust elects a board majority. The offering faces an FTC investigation, questions from Senator Elizabeth Warren, and calls from New York City Comptroller Mark Levine and SOC Investment Group to pause the IPO until its AI safety and governance concerns are addressed.

Calling Plays from the Field: Anthropic Pitches AI Safety Bill

Anthropic representatives privately pitched members of Congress on a draft frontier AI safety and security bill. The bill represents the proposals outlined in Anthropic's Advanced AI Framework and includes language mirroring California's frontier AI laws, such as the Transparency in Frontier Artificial Intelligence Act (SB 53), which was passed in 2025. Notably, the draft bill would:

  • Let the government block a model's release or use only if it poses a foreseeable, imminent risk of more than $1 billion in damages, or of killing or seriously injuring more than 50 people. Such risk must stem from the creation or release of a biological weapon, the model evading its developer's controls, or a cyberattack.
  • Task independent evaluators (which would register with the Department of Commerce) with periodically evaluating frontier labs and their frontier models. If the evaluators determine there is "imminent enumerated risk" for a company's frontier models, they must report it to Commerce within 72 hours of completing their assessment. Labs are permitted to pay the evaluators for their work as long as the payment is not contingent on the evaluations' results.
  • Prohibit any other US department, agency, or officer from suspending the model's development, deployment, or internal use. However, a federal judge for the US District Court for the District of Columbia could order a model to be suspended.
  • Preempt some state AI laws related to public disclosure of risk, third-party evaluation, and reporting of critical incidents. Following criticism, Anthropic has said it will revise the bill's preemption language.
  • Require frontier labs to create safety frameworks, publish transparency reports for new frontier models, implement protections for whistleblowers, and report critical safety incidents to the Department of Commerce.

FTC Orders Airlines, Gig Platforms, and Retailers to Explain Personalized Pricing Algorithms

On October 28, 2026, the Federal Trade Commission issued Section 6(b) orders to nine major airlines, ride-sharing and food-delivery companies, and retailers. The orders require them to explain how they use personal data to set individualized prices for consumers and pay for gig workers. The FTC wants to know whether prices vary from person to person based on factors such as "participation in loyalty programs or analysis of personal data by superintelligence algorithms." The orders reach the methodologies, data inputs, AI and software tools, A/B tests, and consumer disclosures behind these decisions. They also cover internal communications, including content created with generative AI platforms. Recipients have 45 days to respond. The inquiry follows the FTC's August proposed enforcement policy statement. That statement warned that using personal data to set individualized prices without disclosing it could violate the FTC Act.

Awe and Anxiety as AI Races ahead in Mathematics

When OpenAI released hundreds of AI-generated math findings in a single day in October 2026, leading mathematicians reacted with awe: as the New York Times reported, the chair of Rutgers' math department said one proof would have earned "an instant Fields Medal" if a human had produced it, and the president of the Clay Mathematics Institute called the release "breathtaking." That excitement came with real unease, as the Clay Institute president told the Times that the release would be "exciting for many, frightening for others, and devastating for some," and graduate students told the paper that they now question what a math degree will be worth. Mathematicians also found some papers unreadable.

For businesses, the episode raises unsettled legal questions: work produced by AI with little human input may not be protected by copyright, sharing unpublished ideas with AI tools may put them at risk, and errors in AI output make human review essential.

Frontier AI Exposes Decade-Old Flaws at the Core of the Web

On October 21, 2026, Wiz Research disclosed "PHP2Shell," critical vulnerabilities in the core engine of PHP, which by Wiz's estimate underpins roughly 70% of websites, including WordPress, Drupal, and Joomla sites. Rated as high as CVSS 9.2, the flaws let unauthenticated attackers execute code on servers or extract passwords, tokens, and API keys from memory. Given that the defects lie in PHP itself, even well-secured applications are exposed, and every release since version 5.4 has been affected. PHP issued a patch the same day.

What sets PHP2Shell apart is how it was found. The flawed code went unexploited for more than a decade because weaponizing it was "at the edge of what human researchers could practically achieve." Wiz said that a single researcher using frontier models (including Gemini 4 Argon and GPT-5.6 Sol) did in a week what previously would have taken "a world-class team" months. As AI lowers that barrier, vulnerabilities once dismissed as impractical become viable targets.

Beyond the immediate need to patch, including systems run by vendors and hosts, the episode carries a broader lesson: if AI compresses the window between disclosure and exploitation, regulators, litigants, and counterparties may judge the reasonableness of patch-management and incident-response practices against a much shorter clock.

Recent Binance Hack Highlights AI-Era Security Risks

On October 29, 2026, the cryptocurrency exchange Binance lost at least $6.3 billion in a hack that the FBI has attributed to a North Korean group known as TraderTraitor. Binance has reported the hack targeted multiple employees, including one who downloaded malicious software to install an AI coding tool onto a company laptop and another who was instructed to download malware onto a company computer in connection with an "assignment" for a sham job interview. The cryptocurrency analytics firm Chainalysis identified patterns in the flow of funds — particularly, thousands of transactions in a short period, many of which involve repeated circulation of funds among addresses — that suggest automation was used. Although North Korea is known for its sophisticated hacking abilities, experts interviewed by the New York Times opined that AI would make it easier for much less sophisticated hackers to conduct such attacks. While the hack demonstrates ways in which malicious actors can use AI to further exploit companies' vulnerabilities, Binance has said that it is now working with AI companies to better protect its systems and prevent such attacks in the future. Other companies seeking to avoid such attacks may also begin to do the same.

Banks Reportedly Find Unauthorized AI Agents in Customer Accounts and Move to Block Them

Axios reported, according to unnamed sources, that major US banks are finding unauthorized third-party AI agents inside customer accounts, and that some of those agents appear to act without user instructions. In most cases, customers gave the agents their own login credentials; but even with the customers' permission, this can violate bank rules requiring account holders to log in themselves. Several banks are now blocking or restricting such agents, while others are amending their terms of service to shift the risk to customers.

FSB Finalizes Its Sound Practices for the Responsible Adoption of AI

The Financial Stability Board published its final report on the Sound Practices for the Responsible Adoption of Artificial Intelligence. The report identifies 12 sound practices organized into two groups that: (1) emphasize the importance of organization-wide AI governance as financial institutions adopt AI; and (2) set out a comprehensive framework for managing risks throughout the AI lifecycle, from development to deployment. The practices are nonprescriptive and are intended to help financial institutions harness the benefits of AI while managing its risks responsibly. Financial institutions should consider how the practices align with their adoption of AI and the specific types of AI in use, as not all practices will apply to all AI use cases. The report encourages board members and senior management of financial institutions to reference and revisit the practices when developing and refining their AI governance, risk management, and control frameworks as AI adoption expands over time.

The material contained in this communication is informational, general in nature and does not constitute legal advice. The material contained in this communication should not be relied upon or used without consulting a lawyer to consider your specific circumstances. This communication was published on the date specified and may not include any changes in the topics, laws, rules or regulations covered. Receipt of this communication does not establish an attorney-client relationship. In some jurisdictions, this communication may be considered attorney advertising.