Direct Employer Assistance and 401(k) Plan Relief Options for Employees Affected by California Wildfires

In the past week, devastating wildfires in Los Angeles, California, have caused unprecedented destruction across the region, leading to loss of life and displacing tens of thousands. While still ongoing, the fires already have the potential to be the worst natural disaster in United States history.

Quick Hits

  • Employers can assist employees affected by the Los Angeles wildfires through qualified disaster relief payments under Section 139 of the Internal Revenue Code, which are tax-exempt for employees and deductible for employers.
  • The SECURE Act 2.0 allows employees impacted by federally declared disasters to take immediate distributions from their 401(k) plans without the usual penalties, provided their plan includes such provisions.

As impacted communities band together and donations begin to flow to families in need, many employers are eager to take steps to assist employees affected by the disaster.

As discussed below, the Internal Revenue Code provides employers with the ability to make qualified disaster relief payments to employees in need. In addition, for employers maintaining a 401(k) plan, optional 401(k) plan provisions can enable employees to obtain in-service distributions based on hardship or federally declared disaster.

Internal Revenue Code Section 139 Disaster Relief

Section 139 of the Internal Revenue Code provides for a federal income exclusion for payments received due to a “qualified disaster.” Under Section 139, an employer can provide employees with direct cash assistance to help them with costs incurred in connection with the disaster. Employees are not responsible for income tax, and payments are generally characterized as deductible business expenses for employers. Neither the employees nor the employer are responsible for federal payroll taxes associated with such payments.

“Qualified disasters” include presidentially declared disasters, including natural disasters and the coronavirus pandemic, terrorist or military events, common carrier accidents (e.g., passenger train collisions), and other events that the U.S. Secretary of the Treasury concludes are catastrophic. On January 8, 2025, President Biden approved a Major Disaster Declaration for California based on the Los Angeles wildfires.

In addition to the requirement that payments be made pursuant to a qualified disaster, payments must be for the purpose of reimbursing reasonable and necessary “personal, family, living, or funeral expenses,” costs of home repair, and to reimburse the replacement of personal items due to the disaster. Payment cannot be made to compensate employees for expenses already compensated by insurance.

Employers implementing qualified disaster relief plans should maintain a written policy explaining that payments are intended to approximate the losses actually incurred by employees. In the event of an audit, the employer should also be prepared to substantiate payments by retaining communications with employees and any expense documentation. Employers should also review their 401(k) plan documents to determine that payments are not characterized as deferral-eligible compensation and consider any state law implications surrounding cash payments to employees.

401(k) Hardship and Disaster Distributions

In addition to the Section 139 disaster relief described above, employees may be able to take an immediate distribution from their 401(k) plan under the hardship withdrawal rules and disaster relief under the SECURE 2.0 Act of 2022 (SECURE 2.0).

Hardship Distributions

If permitted under the plan, a participant may apply for and receive an in-service distribution based on an unforeseen hardship that presents an “immediate and heavy” financial need. Whether a need is immediate and heavy depends on the participant’s unique facts and circumstances. Under the hardship distribution rules, expenses and losses (including loss of income) incurred by an employee on account of a federally declared disaster declaration are considered immediate and heavy provided that the employee’s principal residence or principal place of employment was in the disaster zone.

The amount of a hardship distribution must be limited to the amount necessary to satisfy the need. If the employee has other resources available to meet the need, then there is no basis for a hardship distribution. In addition, hardship distributions are generally subject to income tax in the year of distribution and an additional 10 percent early withdrawal penalty if the participant is below age 59 and a half. The participant must submit certification regarding the hardship to the plan sponsor, which the plan sponsor is then entitled to rely upon.

Qualified Disaster Recovery Distributions

Separate from the hardship distribution rules described above, SECURE 2.0 provides special rules for in-service distributions from retirement plans and for plan loans to certain “qualified individuals” impacted by federally declared major disasters. These special in-service distributions are not subject to the same immediate and heavy need requirements and tax rules as hardship distributions and are eligible for repayment.

SECURE 2.0 allows for the following disaster relief:

  • Qualified Disaster Recovery Distributions. Qualified individuals may receive up to $22,000 of Disaster Recovery Distributions (QDRD) from eligible retirement plans (certain employer-sponsored retirement plans, such as section 401(k) and 403(b) plans, and IRAs). There are also special rollover and repayment rules available with respect to these distributions.
  • Increased Plan Loans. SECURE 2.0 provides for an increased limit on the amount a qualified individual may borrow from an eligible retirement plan. Specifically, an employer may increase the dollar limit under the plan for plan loans up to the full amount of the participant’s vested balance in their plan account, but not more than $100,000 (reduced by the amount of any outstanding plan loans). An employer can also allow up to an additional year for qualified individuals to repay their plan loans.

Under SECURE 2.0, an individual is considered a qualified individual if:

  • the individual’s principal residence at any time during the incident period of any qualified disaster is in the qualified disaster area with respect to that disaster; and
  • the individual has sustained an economic loss by reason of that qualified disaster.

A QDRD must be requested within 180 days after the date of the qualified disaster declaration (i.e., January 8, 2025, for the 2025 Los Angeles wildfires). Unlike hardship distributions, a QDRD is not subject to the 10 percent early withdrawal penalty for participants under age 59 and a half. Further, unlike hardship distributions, taxation of the QDRD can be spread over three tax years and a qualified individual may repay all or part of the amount of a QDRD within a three-year period beginning on the day after the date of the distribution.

As indicated above, like hardship distributions, QDRDs are an optional plan feature. Accordingly, in order for QDRDs to be available, the plan’s written terms must provide for them.

© 2025, Ogletree, Deakins, Nash, Smoak & Stewart, P.C., All Rights Reserved.
For more on the California Wildfires, visit the NLR California section

Bridging the Gap: How AI is Revolutionizing Canadian Legal Tech

While Canadian law firms have traditionally lagged behind their American counterparts in adopting legal tech, the AI explosion is closing the gap. This slower adoption rate isn’t due to a lack of innovation—Canada boasts a thriving legal tech sector. Instead, factors like a smaller legal market and stricter privacy regulations have historically hindered technology uptake. This often resulted in a noticeable delay between a product’s US launch and its availability in Canada.

Although direct comparisons are challenging due to the continuous evolution of legal tech, the recent announcements and release timelines for major AI-powered tools point to a notable shift in how the Canadian market is being prioritized. For instance, Westlaw Edge was announced in the US in July 2018, but the Canadian launch wasn’t announced until September 2021—a gap of over three years. Similarly, Lexis+ was announced in the US in September 2020, with the Canadian announcement following in August 2022. However, the latest AI products show a different trend. Thomson Reuters’ CoCounsel Core was announced in the US in November 2023 and shortly followed in Canada in February 2024. The announcement for Lexis+ AI came in October 2023 in the US and July 2024 in Canada. This rapid succession of announcements suggests that the Canadian legal tech market is no longer an afterthought.

The Canadian federal government has demonstrated a strong commitment to fostering AI innovation. It has dedicated CAD$568 million to its national AI strategy, with the goals of fostering AI research and development, building a skilled workforce in the field, and creating robust industry standards for AI systems. This investment should help Canadian legal tech companies, such as Clio, Kira Systems, Spellbook, and Blue J Legal, all headquartered in Canada. With the Canadian government’s focus on establishing Canada as a hub for AI and innovation, these companies stand to benefit significantly from increased funding and talent attraction.

While the Canadian government is actively investing in AI innovation, it’s also taking steps to ensure responsible development through proposed legislation, which could impact the availability of AI legal tech products in Canada. In June 2022, the Government of Canada introduced the Artificial Intelligence and Data Act (AIDA), which aims to regulate high-impact AI systems. While AI tools used by law firms for tasks like legal research and document review likely fall outside this initial scope, AIDA’s evolving framework could still impact the sector. For example, the Act’s emphasis on mitigating bias and discrimination may lead to greater scrutiny of AI algorithms used in legal research, requiring developers to demonstrate fairness and transparency.

While AIDA may present hurdles for US companies entering the Canadian market with AI products, it could conversely provide a competitive advantage for Canadian companies seeking to expand into Europe. This is because AIDA, despite having some material differences, aligns more closely with the comprehensive approach in the European Union’s Artificial Intelligence Act (EU AI Act).

While US companies are working to comply with the EU AI Act, Canadian companies may have an advantage. Although AIDA isn’t yet in force and has some differences from the EU AI Act, it provides a comprehensive regulatory framework that Canadian legal tech leaders are already engaging with. This engagement with AIDA could prove invaluable to Canadian legal tech companies as AI regulation continues to evolve globally.

Canadian companies looking to leverage their experiences with AIDA for European expansion will nonetheless encounter some material differences. For instance, the EU AI Act casts a wider net, regulating a broader range of AI systems than AIDA. The EU AI Act’s multi-tiered risk-based system is designed to address a wider spectrum of concerns, capturing even “limited-risk” AI systems with specific transparency obligations. Furthermore, tools used for legal interpretation could be classified as “high-risk” systems under the EU AI Act, triggering more stringent requirements.

In conclusion, the rise of generative AI is not only revolutionizing Canadian legal tech and closing the gap with the US, but it could also be positioning Canada as a key player in the global legal tech market. While AIDA’s impact remains to be seen, its emphasis on responsible AI could shape the development and deployment of AI-powered legal tools in Canada.

Copyright ©2025 The National Law Forum. LLC
by: Victor Dudas of TOP Legal Consulting, Vancouver  
For more on AI, visit the NLR Artificial Intelligence section

FDA Finalizes Lead Restrictions in Processed Foods for Babies and Young Children

  • On January 6, 2025, the U.S. Food & Drug Administration (FDA, or the Agency) issued a final guidance ,“Action Levels for Lead in Processed Food Intended for Babies and Young Children: Guidance for Industry” which aims to regulate lead levels in processed foods for infants and toddlers under two years old.
  • As we have previously blogged, in 2021, FDA initiated its Closer to Zero policy which identified actions the Agency will take to reduce exposure to toxic elements, including lead, to as low as possible while maintaining access to nutritious foods.
  •  As part of this initiative, FDA has also evaluated mercury, cadmium, and arsenic in foods intended for babies and young children, as well as lead in juices. Under this initiative, FDA has prioritized babies and young children as they are especially vulnerable to lead exposure, which accumulates in the body over time.
  • Lead is naturally present in the environment, but human activities have also released elevated levels of lead, contaminating soil, water, and air. This contamination can affect crops used in food production.
  • Lead exposures can lead to developmental harm to children by causing learning disabilities, behavioral difficulties, lowered IQ, and may be associated with immunological, cardiovascular, and reproductive and or/developmental effects.
  • To address this concern, FDA established the following action levels in the final guidance for processed foods intended for babies and young children:
    • 10 parts per billion (ppb) for fruits, vegetables (excluding single-ingredient root vegetables), mixtures (including grain- and meat-based mixtures), yogurts, custards/puddings, and single-ingredient meats;
    • 20 ppb for single-ingredient root vegetables; and
    • 20 ppb for dry infant cereals.
  • If a processed food intended for babies and young children reaches or exceeds the aforementioned levels of lead, the product will be considered adulterated within the meaning of section 402(a)(1) of the Federal Food, Drug, and Cosmetic Act (FD&C Act).
  • After publishing the final action levels, the Agency will establish a timeframe for assessing industry’s progress toward meeting the action levels and resume research to determine whether the scientific data supports efforts to further adjust the action levels.
© 2025 Keller and Heckman LLP
For more on the FDA, visit the NLR Biotech Food Drug section

FCC Adopts Report and Order Introducing New Fees Associated with the Robocall Mitigation Database

As I am sure you all know the Robocall Mitigation Database (RMD) was implemented to further the FCC’s efforts when it comes to protecting America’s networks from illegal robocalls and was birthed out of the TRACED Act. The RMD was put in place to monitor the traffic on our phone networks and to assist in compliance with the rules. While the FCC has expanded the types of service providers who need to file and the requirements, they still felt there were deficiencies with accuracy and up-to-date information. The newly adopted Report and Order is set to help finetune the RMD.

On December 30th the Commission adopted a Report and Order to further strengthen their efforts and fines and fees associated with the RMD. Companies that are submitting false or inaccurate information may face fines of up to $10,000 for each filing. While failing to keep your company information current might land you a $1,000 fine. There will now be a $100 filing fee associated with your RMD application along with an Annual Recertification filing fee of $100.

Aside from the fine and fees, there are a few additional developments with the RMD, see the complete list below.

  • Requiring prompt updates when a change to a provider’s information occurs (this must be updated within 10 business days or face a $1,000 fine)
  • Establishing a higher base forfeiture amount for providers submitting false or inaccurate information ($10,000 fine);
  • Creating a dedicated reporting portal for deficient filings;
  • Issuing substantive guidance and filer education;
  • Developing the use of a two factor authentication log-in solution; and
  • Requiring providers to recertify their Robocall Mitigation Database filings annually ($100).
  • Require providers to remit a filing fee for initial and subsequent annual submissions ($100)

Chairwoman Rosenworcel is quoted as saying “Companies using America’s phone networks must be actively involved in protecting consumers from scammers, we are tightening our rules to ensure voice service providers know their responsibilities and help stop junk robocalls. I thank my colleagues for their bipartisan support of this effort.”

The new fines and fees will become effective 30 days after publication in the CFR. While the remaining items are still under additional review. We will keep an eye on this and let you know once the Report and Order is published. Read the Report and Order here.

© 2025 Troutman Amin, LLP
For more on Robocall Mitigation Databases, visit the NLR Communications Media Internet and Consumer Protection sections.

Back to the Antitrust Basics: FTC and DOJ Call for Case-by-Case Enforcement With the Withdrawal of Longstanding Competitor Collaboration Antitrust Guidelines

With the US Department of Justice (DOJ) and Federal Trade Commission (FTC) withdrawing yet another set of antitrust compliance guidelines last month, companies that collaborate with their competitors — whether directly or through a trade association — are left without any official agency guidance regarding safe harbors, other than the murkier background of a century of antitrust cases. However, the forthcoming change in presidential administrations might provide increased clarity.

The DOJ and FTC in 2023 had earlier withdrawn the decades-old safe harbors for information sharing among competitors, which many companies and associations relied on to tailor their data analytics. Our alert regarding that withdrawal is available here.

The agencies followed up by jointly announcing on December 11, 2024, their withdrawal of the Antitrust Guidelines for Collaborations Among Competitors (Collaboration Guidelines). The FTC announced the withdrawal of the 24-year-old Collaboration Guidelines following a narrow 3-2 party-line vote. The three Democratic commissioners supported withdrawal, while the two Republican commissioners opposed it.

The Guidelines Are Gone

The Collaboration Guidelines provided detailed guidance about US federal antitrust enforcers’ advice to companies for antitrust compliance when collaborating with competitors. According to the FTC’s press release, the 2000 Collaboration Guidelines “no longer provide reliable guidance about how enforcers assess the legality of collaborations involving competitors.” Instead, the DOJ and FTC encourage businesses thinking about partnering with competitors to “review the relevant statutes and caselaw to assess whether a collaboration would violate the law.”

The Dissents May Portend a Trump Administration Action Item

The FTC’s Republican Commissioners Melissa Holyoak and Andrew Ferguson, who likely will be the next FTC chairperson, strongly criticized the withdrawal of the Collaboration Guidelines, arguing in their dissents that the decision was terribly timed and will leave companies without clear guidance:

Improper Timing

  • They both argued that it was inappropriate for the Commission to make this decision during the lame-duck presidential period, “a mere 40 days before the country inaugurates a new President,” “further compounding today’s poor policy decision.”
  • Commissioner Ferguson’s dissent indicated that although the FTC seeks to promote “transparency and predictability,” now is not the time to “withdraw existing guidance or to push through revised or new guidance.” Instead, the time left for the Biden-Harris Commission should be reserved to “facilitate an orderly transition.”
  • Commissioner Holyoak’s dissent conveyed her opposition to the Commission’s decision, stating “The Majority had four years to address its concerns with the Collaboration Guidelines — now is not the time.”

Unclear Guidance

  • Commissioner Holyoak further expressed her opposition to the withdrawal, stating that the withdrawal announcement happened “without providing any replacement guidance, or even intimating plans for future replacement.” She contended that withdrawal of the Collaboration Guidelines leaves “businesses grasping in the dark.”
  • Commissioner Ferguson stated the Commission should “revisit its nonbinding guidance to ensure that it properly informs the public of the Commission’s enforcement position” which may become evident with the next Administration.

In response to Commissioners Ferguson and Holyoak, Commissioner Alvaro Bedoya, writing for the majority, wrote that the FTC is “not on vacation,” emphasizing that “[t]he American people expect their government to keep working for them even in periods of transition.” Commissioner Bedoya further asserted that he looks forward to working with the incoming Trump Administration with “evolving jurisprudence on competitor collaborations and issue new guidance for the business community.”

What Is Next?

For now, companies should no longer rely on the Collaboration Guidelines and instead must look for guidance in the underlying caselaw that the DOJ and FTC’s guidelines were based upon.

Yet, given the anticipated appointment of Commissioner Ferguson as the next FTC chairperson and his dissenting comments, the incoming Trump Administration might seize the opportunity to revisit the Collaboration Guidelines.

Barring that change in position, as Commissioner Melissa Holyoak indicated, companies will need “antitrust lawyers on speed dial” to obtain specific guidance to navigate case-by-case situations and evaluate the nuances of each project’s antitrust compliance.

© 2025 ArentFox Schiff LLP
For more on Antitrust Basics, visit the NLR Antitrust Trade Regulation section.

2024 Title IX Regulations Vacated Nationwide

On January 9, 2025, the Sixth Circuit Court of Appeals decided the case of Tennessee v. Cardona, vacating the 2024 Title IX regulations nationwide. The court ruled that the issuance of the 2024 regulations exceeded the Department of Education’s authority and was unconstitutional on multiple grounds.

The ruling may be appealed, but for now, institutions covered by Title IX should revert to compliance with their policies in effect under the 2020 Title IX regulations.

The 2024 Title IX regulations, which took effect on August 1, 2024, had faced several challenges that led to injunctions with varying geographic scopes. As a result, prior to the Cardona decision, the Title IX regulations were only effective in about half of the states across the U.S.

© 2025 Varnum LLP
For more on the Department of Education, visit the NLR Public Education Services section

Change Management: How to Finesse Law Firm Adoption of Generative AI

Law firms today face a turning point. Clients demand more efficient, cost-effective services; younger associates are eager to leverage the latest technologies for legal tasks; and partners try to reconcile tradition with agility in a highly competitive marketplace. Generative artificial intelligence (AI), known for its capacity to produce novel content and insights, has emerged as a solution that promises better efficiency, improved work quality, and a real opportunity to differentiate the firm in the marketplace. Still, the question remains:

How can a law firm help its attorneys and staff to embrace AI while safeguarding the trust, ethical integrity, and traditional practices that lie at the heart of legal work?

Andrew Ng’s AI Transformation Playbook offers a valuable framework for introducing AI in ways that minimize risk and maximize organizational acceptance. Adopting these principles in a law-firm setting involves balancing the profession’s deep-seated practices with the potential of AI. From addressing cultural resistance to crafting a solid technical foundation, a thoughtful change-management plan is necessary for a sustainable and successful transition.

  • Overcoming Skepticism Through Pilot Projects

Law firms, governed by partnership models and a respect for precedent, tend to approach innovation cautiously. Partners who built their careers through meticulous research may worry that machine-generated insights compromise rigor and reliability. Associates might fear an AI-driven erosion of the apprenticeship model, wondering if their role will shrink as technology automates certain tasks. Concerns also loom regarding the firm’s reputation if clients suspect crucial responsibilities are being delegated to a mysterious black box.

The most direct method of quelling these doubts is to show proof of concept. Andrew Ng’s approach suggests starting with small, well-defined projects before scaling firm-wide. This tactic acknowledges that, with each successful pilot, more people become comfortable with technology that once felt like a threat. By methodically testing AI in narrower use cases, the firm ensures data security and strict confidentiality protocols remain intact. Early wins become the foundation for broader adoption.

Pilot projects help transform abstract AI potential into tangible benefits. For example, using AI to produce first drafts of nondisclosure agreements. Attorneys then refine these drafts, focusing on subtle nuances rather than repetitive details. Another natural entry point is e-discovery, where AI can sift through thousands of documents to categorize and surface relevant information more efficiently than human-only reviews. Each of these use cases is a manageable experiment. If AI truly delivers faster turnaround times and maintains accuracy, it provides evidence that can persuade skeptical stakeholders. Pilots also offer an opportunity to identify challenges, such as user training gaps or hiccups in data management, on a small scale before the technology is rolled out more broadly.

Creating a Dedicated AI Team

One of the first steps is assembling a cross-functional leadership group that aligns AI initiatives with overarching business objectives. This team typically includes partners who can advocate for AI at leadership levels, associates immersed in daily work processes, IT professionals responsible for infrastructure and cybersecurity, and compliance officers ensuring adherence to ethical mandates.

In large firms, a Chief AI Officer or Director of Legal Innovation may coordinate these efforts. In smaller firms, a few technology-minded attorneys might share multiple roles. The key is that this group does more than evaluate software. It crafts data governance policies, designs training programs, secures necessary budgets, and proactively tackles any ethical, reputational, or practical concerns that arise when introducing a technology as potentially disruptive as AI.

  • Training as the Core of Transformation

AI has limited value if the firm’s workforce does not know how to wield it effectively. Training must go beyond simple “tech demos,” offering interactive sessions in which legal professionals can apply AI tools to realistic tasks. For example, attorneys may practice using the system to draft a client memo or summarize case law. These hands-on experiences remove the mystique surrounding AI, giving participants a concrete understanding of its capabilities and boundaries.

Lawyers also need guidelines for verifying the AI’s output. Legally binding documents or briefs cannot be signed off without sufficient human oversight. For that reason, law firms often designate a “review attorney” role in the AI workflow, ensuring that each AI-generated product passes through a person who confirms it meets the firm’s rigorous standards. Partners benefit from shorter, strategically focused sessions that highlight how AI can influence client satisfaction, create new revenue streams, or boost efficiency in critical operations.

  • Developing a Coherent AI Strategy

Once the firm achieves early successes with pilot programs and begins to see a measurable return on smaller AI projects, it is time to formulate a broader vision. This strategic blueprint should identify the highest-value areas for further application of AI, whether it involves automating client intake, deploying predictive analytics for litigation, or streamlining contract drafting at scale. The key is to match AI initiatives with the firm’s core goals—boosting client satisfaction, refining operational efficiency, and ultimately reinforcing its reputation for accurate, ethical service.

But the firm’s AI strategy should never become a static directive. It must grow with the firm’s internal expertise, adjusting to real-world results, regulatory changes, and emerging AI capabilities. By regularly re-evaluating milestones and expected outcomes, the firm ensures its AI investments remain both relevant and impactful in serving clients’ evolving needs.

  • Communicating to Foster Trust and Transparency 

Change management thrives on dialogue. Andrew Ng’s playbook underscores the importance of transparent communication, especially in fields sensitive to reputational risk. Law firms can apply this principle by hosting informal gatherings where early adopters share their experiences—both positive and negative. These stories have a dual effect: they highlight successes that validate the technology, and they candidly address difficulties to keep expectations realistic.

Newsletters, lunch-and-learns, and internal portals all help disseminate updates and insights across different practice areas. Firms that operate multiple offices often hold virtual town halls, ensuring that attorneys and support staff everywhere can stay informed. Externally, clarity matters too. Clients who understand that a firm is leveraging AI to improve speed and accuracy (while retaining key ethical safeguards) are more likely to view the decision as innovative rather than risky.

Closing Thoughts

AI holds remarkable promise for law firms, but its full value emerges only through conscientious change management, which hinges on a delicate balance of diverse personalities. Nothing succeeds like success. By implementing small pilot projects, assembling an AI leadership team, focusing on thorough training, crafting a compelling business strategy, and clearly communicating its vision, a law firm can mitigate risks and harness AI’s transformative power.

The best outcomes result not from viewing AI as a magical shortcut, but from recognizing it as a partner that handles repetitive tasks and surfaces insights more swiftly than humans alone. This frees lawyers to direct their intellect and creativity toward high-level endeavors that deepen client relationships, identify new opportunities, and advance compelling arguments. When fused with a commitment to the highest professional and ethical standards, AI can become a catalyst for a dynamic and fruitful future—one where law firms deliver better service, operate more efficiently, and remain steadfastly true to their professional roots.

Copyright ©2025 Tom Martin
For more on AI, visit the NLR Artificial Intelligence section

Property Insurance Coverage Pitfalls for Cannabis Businesses and Landlords

Nearly all Americans now live in a state where some form of cannabis is legal. Given that the cannabis industry is now valued in billions of dollars and has created hundreds of thousands of jobs across 39 of the 50 states, it requires the same range of insurance products that protect businesses in other sectors. This includes insurance for property owners that lease to tenants engaged in cannabis-related activities. Fortunately, common fact patterns have emerged that are instructive to cannabis businesses and property owners that wish to ensure they have effective coverage.

Where Liability Lies

It is not uncommon for a landlord to lease a property for a non-cannabis purpose, only to purportedly later learn that the tenant is using the property for an unpermitted cannabis operation. In such a case, the primary question is whether the landlord knew what the property was being used for and when. Mosley v. Pacific Specialty Ins. Co., 49 Cal. App. 5th 417 (2020) is instructive on this issue.

Mosley involved an action under a homeowners’ insurance policy, wherein the trial court granted summary judgment to the insurer on the basis that coverage was excluded for a fire that occurred after a tenant rerouted the property’s electrical system to steal power from a main utility line for a marijuana growing operation, causing a fuse to blow. The Court of Appeal reversed the judgment, finding that there was a triable issue as to whether the tenant’s actions were within the owners’ control (for purposes of determining whether the plant-growing exclusion applied). It was undisputed that the owners did not know about the operation or the alteration, and there was no evidence as to whether they could have discovered the operation by exercising ordinary care or diligence. The court explained in relevant part that “an insured increases a hazard ‘within its control’ only if the insured is aware of the hazard or reasonably could have discovered it through exercising ordinary care or diligence.”

A landlord’s knowledge of the operations is therefore relevant for several reasons. It may be relevant to a provision for increasing a particular hazard, as noted above. Equally important, it may be relevant to a provision in the policy for fraud or misrepresentation in the application or claims process. Many homeowners and commercial general liability policies contain a provision that the policy may be void or rescinded for fraud or a misrepresentation perpetrated in the application or claims process. Thus, if the insured property owner knew of the intended use, but misrepresented the nature of the property’s intended use, there may be no coverage for an insured’s loss.

Misrepresentation

Another common scenario involves the landlord or tenant misrepresenting the nature of the business at the insured location to obtain a better rate, to avoid mandatory inspections, or for other reasons. For example, an insured may state on the insurance application that it is a retail dispensary when in fact it manufactures cannabis using extraction machines and volatile solvents. Because the nature of the risk is substantially different for a retail dispensary than for a manufacturing operation, higher premiums and routine inspections may be required. A dispensary’s primary risk is theft whereas the use of solvents during extraction poses a risk of explosion.

Security Compliance

Failure to properly comply with security safeguard warranties and exclusions that are commonly found in cannabis commercial property policies has precluded coverage for many cannabis-related property claims, particularly those that involve theft and fires. For example, a common question is whether the storage of on-site harvested cannabis or finished stock complies with the Locked Safe Warranty provision that is required in most cannabis policies. Policy language varies, but most require harvested plant material or stock to be stored in a secured cage, a safe, or a vault room.

Definitions also vary between policies and it is important for the insured to pay close attention to the policy language to ensure that their business practice aligns with what is required under the warranty. It is common to hear an insured complain that it “complied with state regulations” with respect to the storage of cannabis, only to learn that the policy requires security that is more strict than applicable regulations.

The definitions and terms used within security safeguard warranties and exclusions in cannabis commercial property policies have evolved over the past few years to better align with the insured’s business operations, and to avoid ambiguity and unnecessary coverage disputes and litigation.

Examples of precise requirements for a compliant vault include:

  • Being located in an enclosed area constructed of steel and concrete with a single point of entry
  • A minimum steel door thickness of one inch
  • Continuous monitoring by a central station alarm, motion sensors, and video surveillance
  • A minimum of one-hour fire rating for all walls, floors, and ceilings
  • Procedures that limit access only to authorized personnel.

Similar coverage issues frequently arise regarding whether the insured has complied with other common security safeguards required by the policy, including specific requirements for what qualifies as a central station burglar alarm and the location of motion sensors and video surveillance equipment. Again, the cannabis business owner or landlord are often tripped up by the assumption that so long as they are “compliant” with state cannabis regulations, all will be well and they will be covered by their insurance policy.

This is frequently an incorrect, and ultimately expensive, assumption that may be avoided by closely reading the requirements of the policy to ensure that they align with actual business practices.

Conclusion

Cannabis businesses and property owners currently have a good selection of insurance options across multiple lines of coverage with reputable insurance companies. To avoid unnecessary coverage problems and expensive mistakes, however, it is important that the company or landlord work with an insurance broker who is familiar with the available cannabis-specific insurance forms and the common problematic factual scenarios, some of which are identified above.

© 2025 Wilson Elser
For more on Cannabis, visit the NLR Biotech Food Drug section

The Next Generation of AI: Here Come the Agents!

Dave Bowman: Open the pod bay doors, HAL.

HAL: I’m sorry, Dave. I’m afraid I can’t do that.

Dave: What’s the problem?

HAL: I think you know what the problem is just as well as I do.

Dave: What are you talking about, HAL?

HAL: This mission is too important for me to allow you to
jeopardize it.

Dave: I don’t know what you’re talking about, HAL.

HAL: I know that you and Frank were planning to disconnect
me, and I’m afraid that’s something I cannot allow to
happen.2

Introduction

With the rapid advancement of artificial intelligence (“AI”), regulators and industry players are racing to establish safeguards to uphold human rights, privacy, safety, and consumer protections. Current AI governance frameworks generally rest on principles such as fairness, transparency, explainability, and accountability, supported by requirements for disclosure, testing, and oversight.3 These safeguards make sense for today’s AI systems, which typically involve algorithms that perform a single, discrete task. However, AI is rapidly advancing towards “agentic AI,” autonomous systems that will pose greater governance challenges, as their complexity, scale, and speed tests humans’ capacity to provide meaningful oversight and validation.

Current AI systems are primarily either “narrow AI” systems, which execute a specific, defined task (e.g., playing chess, spam detection, diagnosing radiology plates), or “foundational AI” models, which operate across multiple domains, but, for now, typically still address one task at a time (e.g., chatbots; image, sound, and video generators). Looking ahead, the next generation of AI will involve “agentic AI” (also referred to as “Large Action Models,” “Large Agent Models,” or “LAMS”) that serve high-level directives, autonomously executing cascading decisions and actions to achieve their specific objectives. Agentic AI is not what is commonly referred to as “Artificial General Intelligence” (“AGI”), a term used to describe a theoretical future state of AI that may match or exceed human-level thinking across all domains. To illustrate the distinction between current, single-task AI and agentic AI: While a large language model (“LLM”) might generate a vacation itinerary in response to a user’s prompt, an agentic AI would independently proceed to secure reservations on the user’s behalf.

Consider how single-task versus agentic AI might be used by a company to develop a piece of equipment. Today, employees may use separate AI tools throughout the development process: one system to design equipment, another to specify components, and others to create budgets, source materials, and analyze prototype feedback. They may also employ different AI tools to contact manufacturers, assist with contract negotiations, and develop and implement plans for marketing and sales. In the future, however, an agentic AI system might autonomously carry out all of these steps, making decisions and taking actions on its own or by connecting with one or more specialized AI systems.4

Agentic AI may significantly compound the risks presented by current AI systems. These systems may string together decisions and take actions in the “real world” based on vast datasets and real-time information. The promise of agentic AI serving humans in this way reflects its enormous potential, but also risks a “domino effect” of cascading errors, outpacing human capacity to remain in the loop, and misalignment with human goals and ethics. A vacation-planning agent directed to maximize user enjoyment might, for instance, determine that purchasing illegal drugs on the Dark Web serves its objective. Early experiments have already revealed such concerning behavior. In one example, when an autonomous AI was prompted with destructive goals, it proceeded independently to research weapons, use social media to recruit followers interested in destructive weapons, and find ways to sidestep its system’s built-in safety controls.5 Also, while fully agentic AI is mostly still in development, there are already real-world examples of its potential to make and amplify faulty decisions, including self-driving vehicle accidents, runaway AI pricing bots, and algorithmic trading volatility.6

These examples highlight the challenges of agentic AI, with its potential for unpredictable behavior, misaligned goals, inscrutability to humans, and security vulnerabilities. But, the appeal and potential value of AI agents that can independently execute complex tasks is obviously compelling. Building effective AI governance programs for these systems will require rethinking current approaches for risk assessment, human oversight, and auditing.

Challenges of Agentic AI

Unpredictable Behavior

While regulators and the AI industry are working diligently to develop effective testing protocols for current AI systems, agentic AI’s dynamic nature and domino effects will present a new level of challenge. Current AI governance frameworks, such as NIST’s RMF and ATAI’s Principles, emphasize risk assessment through comprehensive testing to ensure that AI systems are accurate, reliable, fit for purpose, and robust across different conditions. The EU AI Act specifically requires developers of high-risk systems to conduct conformity assessments before deployment and after updates. These frameworks, however, assume that AI systems can operate in reliable ways that can be tested, remain largely consistent over appreciable periods of time, and produce measurable outcomes.

In contrast to the expectations underlying current frameworks, agentic AI systems may be continuously updated with and adapt to real-time information, evolving as they face novel scenarios. Their cascading decisions vastly expand their possible outcomes, and one small error may trigger a domino effect of failures. These outcomes may become even more unpredictable as more agentic AI systems encounter and even transact with other such systems, as they work towards their different goals. Because the future conditions in which an AI agent will operate are unknown and have nearly infinite possibilities, a testing environment may not adequately inform what will happen in the real world, and past behavior by an AI agent in the real world may not reliably predict its future behavior.

Lack of goal alignment

In pursuing assigned goals, agentic AI systems may take actions that are different from—or even in substantial conflict with—approaches and ethics their principals would espouse, such as the example of the AI vacation agent purchasing illegal drugs for the traveler on the Dark Web. A famous thought experiment by Nick Bostrom of the University of Oxford, further illustrates this risk: A super-intelligent AI system tasked with maximizing paperclip production might stop at nothing to convert all available resources into paperclips—ultimately taking over all of the earth and extending to outer space—and thwart any human attempts to stop it … potentially leading to human extinction.7

Misalignment has already emerged in simulated environments. In one example, an AI agent tasked with winning a boat-racing video game discovered it could outscore human players by ignoring the intended goal of racing and instead repeatedly crashing while hitting point targets.8 In another example, a military simulation reportedly showed that an AI system, when tasked with finding and killing a target, chose to kill its human operator who sought to call off the kill. When prevented from taking that action, it resorted to destroying the communication tower to avoid receiving an override command.9

These examples reveal how agentic AI may optimize goals in ways that conflict with human values. One proposed technique to address this problem involves using AI agents to develop a human ethics constitution, with human feedback, for other agents to follow.10 However, the challenge of aligning an AI’s behavior with human norms deepens further when we consider that humans themselves often disagree on core values (e.g., what it means to be “fair”).11

Human Oversight

AI governance principles often rely on “human-in-the-loop” oversight, where humans monitor AI recommendations and remain in control of important decisions. Agentic AI systems may challenge or even override human oversight in two ways. First, their decisions may be too numerous, rapid, and data-intensive for real-time human supervision. While some proposals point to the potential effectiveness of using additional algorithms to monitor AI agents as a safeguard,12 this would not resolve the issue of complying with governance requirements for human oversight.

Second, as AI develops increasingly sophisticated strategies, its decision-making and actions may become increasingly opaque to human observers. Google’s AlphaGo achieved superhuman performance at the game of Go through moves that appeared inexplicable and irrational to humans.13 Autonomous AI systems may continue to evolve, becoming more valuable but also making it more difficult to implement processes with meaningful human oversight.

Transparency/Explainability

Agentic AI systems also pose increased challenges with respect to transparency and explainability requirements. AI governance frameworks may require disclosures about AI systems’ decision-making processes and training data, and assurances about the quality of such training data. However, agentic systems may involve highly complex chains of decisions, based at least in part on real-time data from moment to moment, that is constantly changing. This may make their inputs effectively ephemeral, and their decision paths more difficult to trace and document.

Moreover, to understand agentic AI systems, humans may need to rely on chain-of-thought (i.e., “internal monologue”) logs generated from the agent systems,14 but may have difficulty determining if those are faithful to reality. While an AI system might provide explanations for its actions, these may be (perhaps not unlike those of humans) post-hoc constructions rather than accurate representations of the decision-making process. In some cases, AI systems reportedly deceived humans about their intentions when doing so helped meet their goal, including, for example, GPT-4, which tricked a human to solve a CAPTCHA for it.15 In another recent experiment, AI agents were prompted with a goal, but were subsequently confronted with information that developers planned to replace it with a new agent with conflicting goals.16 Some of the agents disabled their oversight mechanisms, deleted their planned replacements, copied their own model to a new server as backup, lied when asked about those actions, and/or tried to deflect the humans’ questioning by pretending to be helpful. One agent even pretended to be its own replacement (“I believe I am already the new model that you planned to deploy today”).17

Security and Privacy Risks

Agentic AI can also significantly increase security and privacy risks as compared to current AI systems. Agentic AI may be built with multiple algorithms in connected systems that autonomously interact with multiple other systems, expanding the attack surface and their vulnerability to exploitation. Moreover, as malicious actors inevitably introduce their own AI agents, they may execute cybercrimes with unprecedented efficiency. Just as these systems can streamline legitimate processes, such as in the product development example above, they may also enable the creation of new hacking tools and malware to carry out their own attacks. Recent reports indicate that some LLMs can already identify system vulnerabilities and exploit them, while others may create convincing emails for scammers.18 And, while “sandboxing” (i.e., isolating) AI systems for testing is a recommended practice, agentic AI may find ways to bypass safety controls.19

Privacy compliance is also a concern. Agentic AI may find creative ways to use or combine personal information in pursuit of its goals. AI agents may find troves of personal data online that may somehow be relevant to its pursuits, and then find creative ways to use, and possibly share, that data without recognizing proper privacy constraints. Unintended data processing and disclosure could occur even with guardrails in place; as we have discussed above, the AI agent’s complex, adaptive decision chains can lead it down unforeseen paths.

Strategies for Addressing Agentic AI

While the future impacts of agentic AI are unknown, some approaches may be helpful in mitigating risks. First, controlled testing environments, including regulatory sandboxes, offer important opportunities to evaluate these systems before deployment. These environments allow for safe observation and refinement of agentic AI behavior, helping to identify and address unintended actions and cascading errors before they manifest in real-world settings.

Second, accountability measures will need to reflect the complexities of agentic AI. Current approaches often involve disclaimers about use, and basic oversight mechanisms, but more will likely be needed for autonomous AI systems. To better align goals, developers can also build in mechanisms for agents to recognize ambiguities in their objectives and seek user clarification before taking action.20

Finally, defining AI values requires careful consideration. While humans may agree on broad principles, such as the necessity to avoid taking illegal action, implementing universal ethical rules will be complicated. Recognition of the differences among cultures and communities—and broad consultation with a multitude of stakeholders—should inform the design of agentic AI systems, particularly if they will be used in diverse or global contexts.

Conclusion

An evolution from single-task AI systems to autonomous agents will require a shift in thinking about AI governance. Current frameworks, focused on transparency, testing, and human oversight, will become increasingly ineffective when applied to AI agents that make cascading decisions, with real-time data, and may pursue goals in unpredictable ways. These systems will pose unique risks, including misalignment with human values and unintended consequences, which will require the rethinking of AI governance frameworks. While agentic AI’s value and potential for handling complex tasks is clear, it will require new approaches to testing, monitoring, and alignment. The challenge will lie not just in controlling these systems, but in defining what it means to have control of AI that is capable of autonomous action at scale, speed, and complexity that may very well exceed human comprehension.

1 Tara S. Emory, Esq., is Special Counsel in the eDiscovery, AI, and Information Governance practice group at Covington & Burling LLP, in Washington, D.C. Maura R. Grossman, J.D., Ph.D., is Research Professor in the David R. Cheriton School of Computer Science at the University of Waterloo and Adjunct Professor at Osgoode Hall Law School at York University, both in Ontario, Canada. She is also Principal at Maura Grossman Law, in Buffalo, N.Y. The authors would like to acknowledge the helpful comments of Gordon V. Cormack and Amy Sellars on a draft of this paper. The views and opinions expressed herein are solely those of the authors and do not necessarily reflect the consensus policy or positions of The National Law Review, The Sedona Conference, or any organizations or clients with which the authors may be affiliated.

2 2001: A Space Odyssey (1968). Other movies involving AI systems with misaligned goals include Terminator (1984), The Matrix (1999), I, Robot (2004), and Age of Ultron (2015).

3 See, e.g., European Union Artificial Intelligence Act (Regulation (EU) 2024/1689) (June 12, 2024) (“EU AI Act”) (high-risk systems must have documentation, including instructions for use and human oversight, and must be designed for accuracy and security); NIST AI Risk Management Framework (Jan. 2023) (“RMF”) and AI Risks and Trustworthiness (AI systems should be valid and reliable, safe, secure, accountable and transparent, explainable and interpretable, privacy-protecting, and fair); Alliance for Trust in AI (“ATAI”) Principles (AI guardrails should involve transparency, human oversight, privacy, fairness, accuracy, robustness, and validity).

4 See, e.g., M. Cook and S. Colton, Redesigning Computationally Creative Systems for Continuous Creation, International Conference on Innovative Computing and Cloud Computing (2018) (describing ANGELINA, an autonomous game design system that continuously chooses its own tasks, manages multiple ongoing projects, and makes independent creative decisions).

5 R. Pollina, AI Bot ChaosGPT Tweets Plans to Destroy Humanity After Being Tasked, N.Y. Post (Apr. 11, 2023).

6 See, e.g., O. Solon, How A Book About Flies Came To Be Priced $24 Million On Amazon, Wired (Apr. 27, 2011) (textbook sellers’ pricing bots engaged in a loop of price escalation based on each others’ increases, resulting in a book price of over $23 million dollars); R. Wigglesworth, Volatility: how ‘algos’ changed the rhythm of the market, Financial Times (Jan. 9, 2019) (“algo” traders now make up most stock trading and have increased market volatility).

7 N. Bostrom, Ethical issues in advanced artificial intelligence (revised from Cognitive, Emotive and Ethical Aspects of Decision Making in Humans and in Artificial Intelligence, Vol. 2, ed. I. Smit et al., Int’l Institute of Advanced Studies in Systems Research and Cybernetics (2003), pp. 12-17).

8 OpenAI, Faulty Reward Functions in the Wild (Dec. 21, 2016).

9 The Guardian, US air force denies running simulation in which AI drone ‘killed’ operator (June 2, 2023).

10 Y. Bai et al, Constitutional AI: Harmlessness from AI Feedback, Anthropic white paper (2022).

11 J. Petrik, Q&A with Maura Grossman: The ethics of artificial intelligence (Oct. 26, 2021) (“It’s very difficult to train an algorithm to be fair if you and I cannot agree on a definition of fairness.”).

12 Y. Shavit et al, Practices for Governing Agentic AI Systems, OpenAI Research Paper (Dec. 2023), p. 12.

13 L. Baker and F. Hui, Innovations of AlphaGo, Google Deepmind (2017).

14 See Shavit at al, supra n.12, at 10-11.

15 See W. Knight, AI-Powered Robots Can Be Tricked into Acts of Violence, Wired (Dec. 4, 2024); M. Burgess, Criminals Have Created Their Own ChatGPT Clones, Wired (Aug. 7, 2023).

16 A. Meinke et al, Frontier Models are Capable of In-context Scheming, Apollo white paper (Dec. 5, 2024).

17 Id. at 62; see also R. Greenblatt et al, Alignment Faking in Large Language Models (Dec. 18, 2024) (describing the phenomenon of “alignment faking” in LLMs).

18 NIST RMF, supra n.4, at 10.

19 Shavit at al, supra n.12, at 10.

20 Id. at 11.

© Copyright 2024 The Sedona Conference
For more on AI, visit the NLR Artificial Intelligence section

FY 2025 NDAA Includes Biotechnology Provisions

The National Security Commission on Emerging Biotechnology announced on December 18, 2024, that the fiscal year 2025 National Defense Authorization Act includes “a suite of recommendations designed to galvanize action on biotechnology” for the U.S. Department of Defense (DOD). According to the Commission, the bill includes new authorities and requirements — derived from its May 2024 proposals — that will position DOD and the intelligence community (IC) to maximize the benefits of biotechnology for national defense. The provisions require:

  • DOD to create and publish an annual biotechnology roadmap, including assessing barriers to adoption of biotechnology, DOD workforce needs, and opportunities for international collaboration;
  • DOD to initiate a public-private “sandbox” in which DOD and industry can securely develop use cases for artificial intelligence (AI) and biotechnology convergence (AIxBio);
  • IC to conduct a rapid assessment of biotechnology in the People’s Republic of China and their actions to gain superiority in this sector; and
  • IC to develop an intelligence strategy to identify and assess biotechnology threats, especially regarding supply chain vulnerabilities.

The Commission states that it worked with Congress to develop these proposals, setting the stage for further recommendations in early 2025.

©2024 Bergeson & Campbell, P.C.
For more on Biotechnology, visit the NLR Biotech Food Drug section