Startup Legal Trends for 2026: Emerging Compliance and IP Frameworks

Startup legal trends for 2026 span AI liability, visa rules, patent risk, and Section 1202 stock exclusions every founder and counsel must track now.

Startups in 2026 face ten interlocking compliance and intellectual property pressure points, and the single common thread is timing: Colorado’s repeal of its original AI statute, the vacated then appealed one hundred thousand dollar H 1B fee, the USPTO’s reversed AI inventorship guidance, Delaware’s post Moelis stockholder agreement fixes, and the expanded Internal Revenue Code Section 1202 stock exclusion all reward companies that update their legal documentation continuously and penalize companies that treat compliance as a one time project completed at formation.

Executive Summary

This report examines ten interlocking legal risk areas that determine whether a venture backed technology company survives its next financing round, product launch, or eventual exit without a costly legal surprise. The areas span artificial intelligence liability, international hiring, enterprise contracting, intellectual property strategy for AI systems, data privacy, venture financing diligence, patent clearance and enforcement, worker classification, and exit stage tax and fiduciary duty planning. Regulators and courts moved faster across nearly every one of these areas between 2025 and 2026 than in any comparable prior period, and the direction of that movement is inconsistent: some frameworks are tightening, such as California’s automated decision making rules, while others are retreating, such as Colorado’s repealed AI statute and the European Union’s delayed AI Act deadline.

On artificial intelligence liability, Colorado replaced its comprehensive 2024 AI statute with a narrower disclosure framework in May 2026, while the Federal Trade Commission continued bringing enforcement actions against AI companies under existing consumer protection authority that requires no new statute at all. On international hiring, a federal district court vacated the contested one hundred thousand dollar H 1B fee in June 2026, but the ruling remains under appeal, and a new wage weighted lottery has made the visa pathway structurally harder for junior technical hires regardless of the fee’s ultimate fate. On intellectual property, the United States Patent and Trademark Office loosened its own AI inventorship scrutiny in November 2025 even as the Federal Circuit tightened patent eligibility standards for machine learning claims, creating a gap between how easily a startup can obtain an AI related patent and how well that patent will hold up once challenged.

On venture capital due diligence, Delaware’s response to the Court of Chancery’s 2024 Moelis decision reshaped how board designation rights in stockholder agreements must be documented, while intellectual property assignment gaps remain the most common deal killer identified in Series B and C diligence. On worker classification, the regulatory ground shifted twice in roughly two years: the Department of Labor’s 2024 independent contractor rule went effectively unenforced by mid 2025 and faces formal rescission as of February 2026, while California’s stricter state law standard never moved at all, leaving multi state companies to build compliance around the strictest applicable test rather than the shifting federal one.

The overall picture for 2026 is one of accelerating regulatory and doctrinal change layered onto already complex transactional practice. Startups that treat each of these ten areas as a static compliance checkbox, reviewed once and filed away, are consistently the companies that discover a costly gap at the worst possible moment: during a financing round, a product launch, or an acquisition. Startups that build continuous review cycles into their legal operating rhythm are positioned to treat each regulatory shift as a manageable update rather than a crisis.

Detailed Findings

Subtopic 1: AI Governance and Liability

Link

Artificial intelligence liability regulation moved in a markedly inconsistent direction across 2025 and 2026, with Colorado repealing its own comprehensive AI statute through Senate Bill 26 189, signed May 14, 2026, in favor of a narrower automated decision making disclosure framework effective January 1, 2027, while the European Union’s AI Act high risk provisions, originally set for August 2, 2026, were pushed into a later compliance window following a Digital Omnibus simplification package approved by the Council in June 2026. A United States startup with no European entity can still fall within the AI Act’s extraterritorial reach if its model or interface ends up embedded in a product marketed into the European Union.

Enforcement under existing consumer protection law has continued without pause even as new AI specific statutes stall. The Federal Trade Commission’s Operation AI Comply, proceeding under Section 5 of the FTC Act since September 2024, produced a settlement against DoNotPay over its robot lawyer marketing claims and a March 2026 judgment against Air AI including an eighteen million dollar order, most of it suspended based on documented inability to pay. Neither case required a new statute, and both demonstrate that overstated AI marketing claims remain the highest risk category regardless of what any pending AI bill eventually does.

The National Institute of Standards and Technology’s Generative AI Profile, published July 26, 2024, remains the most stable compliance reference available precisely because it is voluntary rather than tied to any single legislature’s shifting calendar. Startups that build internal AI testing and disclosure documentation around this framework create a defensible record regardless of which state or European deadline eventually takes effect, and increasingly satisfy venture investor diligence expectations that have started treating AI risk documentation as a standard closing condition.

Subtopic 2: International Immigration and Talent Compliance

Link

International hiring compliance in 2026 centers on three unresolved sources of legal uncertainty operating simultaneously. A United States district court in Massachusetts vacated the one hundred thousand dollar per petition H 1B fee on June 8, 2026, but the government’s appeal to the First Circuit remains pending, and a conflicting ruling upholding the same fee out of the District of Columbia means the fee’s ultimate status will not be settled until the appellate process concludes.

Independent of the fee litigation, the Department of Homeland Security’s wage weighted H 1B selection rule, effective February 27, 2026, restructured the lottery itself to favor higher wage tier registrations, making the pathway measurably harder for the entry level and recent graduate hires many startups have historically sponsored. United States Citizenship and Immigration Services separately expanded O 1A extraordinary ability evidentiary standards in January 2025 to credit open source contributions and technical publication in artificial intelligence and emerging technology fields, giving startups a more viable lottery free alternative for senior technical talent.

Visa bulletin backlogs compound both issues for long term retention planning. The July 2026 bulletin shows the employment based second preference category for India unavailable for the remainder of the fiscal year and the first preference category retrogressed, meaning a founder who successfully sponsors an India born employee today is not resolving that employee’s long term status question, only beginning a process that current backlog conditions may not resolve within the employee’s entire tenure at the company.

Subtopic 3: Enterprise Contracting

Link

Enterprise software contracting in 2026 centers on a widening gap between standard liability caps and actual breach exposure, with average data breach costs exceeding four and a half million dollars against contract fee based caps that frequently leave the overwhelming majority of that exposure uncovered absent specific carve outs for data breach and intellectual property indemnification claims. Enterprise procurement teams have begun running this arithmetic before signing rather than after, making the carve out negotiation a standard feature of any enterprise deal rather than an unusual customer demand.

Artificial intelligence features embedded inside SaaS products have become a distinct negotiation category in their own right. Enterprise data processing addenda increasingly require disclosure of every subprocessor including any large language model provider integrated into the product, and enterprise buyers now routinely demand explicit contractual assurance that customer data will not train or fine tune any underlying model absent affirmative opt in consent, a materially stricter standard than the opt out defaults common in earlier contract generations.

Service organization control audit scope has expanded correspondingly. The American Institute of Certified Public Accountants updated its trust services criteria to address AI driven systems, and enterprise security reviews in 2026 expect a SOC 2 Type II report to document model lineage, inference logging, and third party model vendor risk assessment, meaning a startup’s audit completed before these criteria existed likely requires a materially expanded scope before its next enterprise renewal cycle.

Subtopic 4: IP Strategy for AI Generated Assets

Link

The United States Patent and Trademark Office reversed its own artificial intelligence inventorship guidance in November 2025, less than two years after adopting a February 2024 framework that scrutinized whether a human made a significant contribution to an AI assisted invention. The revised posture presumes valid human inventorship whenever any natural person is named, a shift observers have characterized as a don’t ask, don’t tell examination policy that measurably eases the application process.

That relative ease at the examination stage sits in direct tension with an increasingly skeptical Federal Circuit validity posture. Recentive Analytics v. Fox Corp., decided April 2025, held that applying an off the shelf machine learning model to a new data domain without a claimed technical improvement to the model itself produces an unpatentable abstract idea under the Alice framework, continuing a pattern of decisions that make many AI related patents considerably more vulnerable to invalidity challenge than their examination history would suggest.

Trade secret protection under the Defend Trade Secrets Act offers a materially different risk profile for the training methodology and model architecture assets a patent strategy struggles to protect reliably. OpenEvidence v. Pathway Medical, filed in the District of Massachusetts in February 2025, is the first significant test of whether an AI system’s prompts and hidden instructions qualify as trade secrets, and its resolution will shape how AI companies structure access to their own deployed systems going forward.

Subtopic 5: Data Privacy and Security Frameworks

Link

Data privacy compliance obligations expanded on multiple fronts simultaneously heading into 2026. The Federal Trade Commission’s amended Children’s Online Privacy Protection Rule, published April 22, 2025, requires full compliance by April 22, 2026 and expands the definition of personal information to include biometric identifiers while requiring separate verifiable parental consent for third party disclosure used in targeted advertising, the first substantive update to the rule in more than a decade.

California’s Privacy Protection Agency approved automated decision making technology regulations in September 2025 that impose staggered obligations, with risk assessment requirements beginning January 1, 2026 and the core notice and opt out rights for significant decisions in employment, lending, and healthcare contexts taking effect January 1, 2027. Indiana, Kentucky, and Rhode Island each brought new comprehensive consumer privacy statutes into effect January 1, 2026, bringing the total number of states with comprehensive privacy statutes to twenty and each carrying its own thresholds, cure periods, and penalty structures.

European enforcement activity underscores that cross border exposure has not diminished even as domestic United States rulemaking absorbs most attention. Cumulative fines issued under the General Data Protection Regulation since 2018 now exceed seven point one billion euros, including a five hundred thirty million euro fine against a major platform in May 2025 over unlawful data transfers, and any startup with even a modest European user base can fall within that enforcement authority regardless of whether it ever established a European entity.

Subtopic 6: Venture Capital Due Diligence Prep

Link

Series B and C due diligence in 2026 centers on two recurring failure points: incomplete intellectual property assignment chains, particularly for founder created technology predating incorporation and for early contractor work performed before formal legal review existed, and stale board authority documents drafted before the Delaware Court of Chancery’s 2024 Moelis decision.

Board composition has become an independent diligence and drafting risk following the Delaware Court of Chancery’s February 2024 decision in West Palm Beach Firefighters’ Pension Fund v. Moelis & Co., which held that several board composition and preapproval provisions in a stockholder agreement were facially invalid under Section 141(a) of the Delaware General Corporation Law because they impermissibly constrained the board’s statutory authority. Counsel preparing a company for Series B or C diligence must confirm that any existing stockholder agreement or side letter granting board designation rights was drafted, or has been amended, to survive the post Moelis standard.

Standard VC due diligence request lists, exemplified by publicly available templates such as Cooley GO’s sample due diligence request list and the National Venture Capital Association’s model legal documents, updated October 2025, are designed to surface exactly these patterns. A company that has not maintained a living data room mapped to these categories will spend the first several weeks of diligence locating documents rather than answering substantive questions.

Subtopic 7: Freedom to Operate Analysis

Link

A freedom to operate analysis assesses whether a company’s product or feature is likely to infringe a third party’s active patent rights, and the stakes for skipping one before a feature launch have risen sharply. Nonpracticing entities accounted for approximately ninety percent of all patent litigation in the high technology sector during 2025, and nine of the ten most active patent plaintiffs that year build no products of their own.

Cost and timeline have historically been a barrier: traditional attorney led freedom to operate investigations have run from twenty thousand to fifty thousand dollars for a deep, multi jurisdiction investigation, though artificial intelligence assisted search tools are now compressing both cost and timeline for the first pass triage stage while remaining a poor substitute for attorney judgment on claim construction and infringement analysis.

Emerging risk concentrates specifically in agentic orchestration, the methods by which AI agents plan, evaluate, and execute multi step tasks, an area large technology companies have patented aggressively. AI.Law’s June 2026 suit against Eve Legal and its parent Butler Labs, alleging infringement of patents covering AI assisted document drafting, illustrates the litigation exposure directly, with the average cost of defending a complex AI patent infringement suit now exceeding three and a half million dollars.

Subtopic 8: Worker Classification and HR Policies

Link

Worker classification law has been unusually unstable across 2024 through 2026. The Department of Labor’s January 2024 independent contractor rule, effective March 11, 2024 and codified at 29 C.F.R. Part 795, went substantively unenforced by the Wage and Hour Division starting May 1, 2025, and the Department published a formal Notice of Proposed Rulemaking on February 26, 2026 seeking to rescind the rule entirely in favor of a modified 2021 framework.

California imposes an independently codified and stricter standard regardless of federal rulemaking. Under Labor Code Section 2775 and the underlying ABC test drawn from S.G. Borello & Sons, Inc. v. Department of Industrial Relations, a worker is presumed an employee unless the hiring company satisfies all three prongs of the test, a standard that has not moved regardless of the federal government’s repeated reversals.

Enforcement exposure remains severe and is not limited to gig economy platforms. California civil penalties for willful misclassification can reach substantial per violation amounts, and multi state startups face a compounding compliance problem because state level standards vary considerably, making a national policy set at the strictest applicable standard the more defensible, if more expensive, approach for growth stage companies operating across jurisdictions.

Subtopic 9: Automated Patent Enforcement

Link

Startups with patent portfolios increasingly rely on artificial intelligence powered monitoring platforms to detect infringement and generate claim charts in hours rather than months, a capability that arrives at a moment when the underlying litigation environment has grown considerably more aggressive. Nonpracticing entities added 2,236 new defendants to patent litigation campaigns in 2025, an increase of 18.7 percent over 2024, with the technology sector absorbing 44.8 percent of that volume.

Startups issuing their own cease and desist letters face a jurisdictional trap clarified by the Federal Circuit in Jack Henry & Associates, Inc. v. Plano Encryption Technologies, LLC, which held that sending a cease and desist letter into a forum can itself establish sufficient minimum contacts to subject the sender to personal jurisdiction and a declaratory judgment suit in the recipient’s chosen court, a risk that automated or templated enforcement workflows can trigger inadvertently across a large volume of letters.

Pending legislation, the RESTORE Patent Rights Act of 2025, would restore a rebuttable presumption of injunctive relief for prevailing patent holders, a change that would materially increase settlement pressure available to legitimate operating companies and nonpracticing entities alike, raising the stakes for getting enforcement workflow and jurisdictional strategy right before any such change takes effect.

Subtopic 10: Corporate Exit Readiness and Equity Structuring

Link

Tax efficient equity structuring through an eventual acquisition or public offering now hinges on three interlocking rules. Section 83(b) elections remain the single most consequential and most frequently missed founder level filing, requiring election within thirty days of an equity grant to lock in ordinary income recognition at the low valuation typical of an early stage grant rather than a materially higher value at vesting.

Internal Revenue Code Section 1202, as amended by the One Big Beautiful Bill Act signed July 4, 2025, expanded the Qualified Small Business Stock gain exclusion meaningfully, introducing a tiered holding period structure, raising the qualifying gross asset threshold to seventy five million dollars, and increasing the per issuer exclusion cap to fifteen million dollars, changes that materially affect entity structuring and equity grant timing decisions for companies approaching that asset threshold.

On the exit side, the Delaware Corwin cleansing doctrine continues to be the dominant tool for insulating negotiated mergers from post closing fiduciary duty litigation. The Delaware Supreme Court’s February 3, 2025 en banc affirmance in the Anaplan merger litigation confirmed that a fully informed, uncoerced stockholder vote triggers business judgment protection even where the final price reflects a significant reduction from the originally announced deal terms, making disclosure quality, not deal price, the primary determinant of post closing litigation exposure.

Cross Cutting Themes

Documentation timing is the connective thread across all ten subtopics: the thirty day Section 83(b) deadline, the post Moelis stockholder agreement drafting window many firms are now using to review legacy board provisions, the recency requirements for a defensible freedom to operate opinion, and the pre use AI model testing documentation regulators increasingly expect all penalize delay in the same structural way. Delaware’s ongoing statutory and judicial recalibration, from the DGCL amendments following Moelis to the continued strength of Corwin, means board authority and exit documents drafted even two years ago may already be out of step with current doctrine. Federal regulatory instability, visible in the Department of Labor’s reversal on independent contractor enforcement and in Colorado’s repeal of its own AI statute, means startups cannot rely on any single rule remaining fixed and must build compliance programs against the strictest currently applicable standard rather than the most convenient one. International hiring compounds this instability: the wage weighted H 1B lottery and the contested fee mean a company’s talent strategy now depends on litigation outcomes it does not control. Finally, artificial intelligence is compressing cost and timeline in freedom to operate analysis and patent enforcement monitoring alike, but in both directions, lowering the cost of defensive clearance while simultaneously lowering the cost of assertion for nonpracticing entities, raising the aggregate volume of patent risk every startup must manage regardless of its own technology choices.

Open Questions

Whether the First Circuit will uphold or reverse the June 2026 district court ruling vacating the one hundred thousand dollar H 1B fee, a decision that will determine whether startups must budget for the fee as a fixed cost of technical sponsorship or treat it as a resolved, no longer applicable risk.

Whether Colorado’s twice delayed automated decision making disclosure framework, now set for January 1, 2027, will actually take effect on that date or face a third postponement, and whether other states will adopt similar disclosure only models rather than the comprehensive duty of care approach Colorado abandoned.

Whether the Delaware DGCL amendments responding to Moelis will be tested in new Court of Chancery litigation before the current legislative fix is treated as settled, leaving open how broadly courts will read the amendments’ safe harbor.

Whether the Department of Labor’s February 2026 proposed rule rescinding the 2024 independent contractor standard will be finalized before a change in political leadership reopens the question yet again.

Whether the RESTORE Patent Rights Act of 2025 will pass and restore a presumption of injunctive relief, which would materially increase settlement pressure in every freedom to operate and patent enforcement decision startups currently face.

How aggressively the Federal Circuit’s personal jurisdiction holding in Jack Henry v. Plano Encryption Technologies will be applied to templated or automated cease and desist correspondence generated by patent monitoring platforms.

Whether the expanded Section 1202 gross asset threshold of seventy five million dollars will materially change equity grant timing and entity structuring advice for companies approaching that ceiling before a future financing round.

Source List

Colorado Senate Bill 26 189, signed May 14, 2026; repeals and replaces the original Colorado AI Act with a narrower automated decision making disclosure framework effective January 1, 2027.

European Union AI Act, high risk system provisions and the June 2026 Digital Omnibus simplification package; governs extraterritorial application to United States based model and API providers.

Federal Trade Commission, Operation AI Comply enforcement actions including FTC v. DoNotPay and FTC v. Air AI; establishes Section 5 FTC Act liability for overstated AI marketing claims.

National Institute of Standards and Technology, AI Risk Management Framework Generative AI Profile, NIST AI 600 1, published July 26, 2024; the leading voluntary technical risk documentation standard.

Department of Homeland Security, Weighted Selection Process for Registrants and Petitioners Seeking to File Cap Subject H 1B Petitions, effective February 27, 2026; governs the fiscal year 2027 H 1B lottery.

State of California, et al. v. Mullin, et al., U.S. District Court for the District of Massachusetts, decided June 8, 2026; vacated the one hundred thousand dollar H 1B fee proclamation, currently on appeal.

United States Citizenship and Immigration Services, Policy Manual Volume 2, Part M, Policy Alert PA 2025 02, January 8, 2025; expands O 1A extraordinary ability evidentiary standards for AI and technical talent.

United States Patent and Trademark Office, Inventorship Guidance for AI Assisted Inventions, February 2024, and its November 2025 revision; governs whether AI assisted inventions can be patented and by whom.

Thaler v. Vidal, 43 F.4th 1207 (Fed. Cir. 2022); holds that a patent inventor must be a natural person.

Recentive Analytics v. Fox Corp., Federal Circuit, April 2025; applies Alice Section 101 analysis to hold generic machine learning applications unpatentable absent a claimed technical improvement.

Defend Trade Secrets Act, 18 U.S.C. Section 1836, as applied in OpenEvidence v. Pathway Medical Inc., U.S. District Court for the District of Massachusetts, filed February 2025; tests whether AI model prompts qualify as trade secrets.

California Privacy Protection Agency, automated decision making technology regulations, approved September 2025; staggered compliance beginning January 1, 2026 and January 1, 2027.

Federal Trade Commission, amended Children’s Online Privacy Protection Rule, 16 C.F.R. Part 312, published April 22, 2025; requires full compliance by April 22, 2026.

Delaware Court of Chancery, West Palm Beach Firefighters’ Pension Fund v. Moelis & Co., decided February 2024; the foundational board composition and stockholder agreement decision reshaping venture financing documentation.

Delaware General Corporation Law Section 141(a) and subsequent legislative amendments; the controlling statutory authority for board management following the Moelis decision.

U.S. Department of Labor, Employee or Independent Contractor Classification Under the Fair Labor Standards Act, 29 C.F.R. Part 795, effective March 11, 2024, and the February 26, 2026 Notice of Proposed Rulemaking seeking its rescission.

California Labor Code Section 2775 and S.G. Borello & Sons, Inc. v. Department of Industrial Relations, 48 Cal. 3d 341 (1989); the controlling California worker classification standard independent of federal rulemaking.

Internal Revenue Code Section 1202, as amended by the One Big Beautiful Bill Act, Public Law 119 21, signed July 4, 2025; governs the Qualified Small Business Stock gain exclusion.

Internal Revenue Code Section 83(b); governs founder equity tax elections and the thirty day filing deadline.

Delaware Supreme Court, en banc affirmance of Court of Chancery Corwin cleansing analysis in the Anaplan merger litigation, February 3, 2025; the controlling post closing fiduciary duty standard for negotiated mergers.

Federal Circuit, Jack Henry & Associates, Inc. v. Plano Encryption Technologies, LLC; establishes personal jurisdiction risk for senders of cease and desist patent correspondence.

AI.Law v. Eve Legal and Butler Labs Inc., U.S. District Court for the Northern District of California, filed June 17, 2026; illustrates current patent litigation exposure for AI assisted document generation products.

National Venture Capital Association, Model Legal Documents, updated October 2025; the industry standard financing and diligence documentation benchmark.

IPWatchdog, U.S. Patent Litigation Trends in 2025: Patterns Behind the Numbers, September 28, 2025; market data on nonpracticing entity litigation volume and technology sector concentration.

Bibliography

AI.Law v. Eve Legal and Butler Labs Inc. U.S. District Court for the Northern District of California, filed June 17, 2026.

Baker Tilly. “Changes to Section 1202, Qualified Small Business Stock, in the One Big Beautiful Bill Act.” 2025.

California Labor Code Section 2775.

California Privacy Protection Agency. Automated Decision Making Technology Regulations. Approved September 2025.

Colorado General Assembly. Senate Bill 26 189. Signed May 14, 2026.

Delaware Court of Chancery. West Palm Beach Firefighters’ Pension Fund v. Moelis & Co. February 2024.

Delaware General Corporation Law Section 141.

Delaware Supreme Court. Anaplan merger litigation, en banc affirmance. February 3, 2025.

Federal Circuit. Jack Henry & Associates, Inc. v. Plano Encryption Technologies, LLC.

Federal Circuit. Recentive Analytics v. Fox Corp. April 2025.

Federal Circuit. Thaler v. Vidal. 43 F.4th 1207. 2022.

Federal Trade Commission. Operation AI Comply enforcement actions, including FTC v. DoNotPay and FTC v. Air AI. 2024 through 2026.

Federal Trade Commission. Amended Children’s Online Privacy Protection Rule. 16 C.F.R. Part 312. April 22, 2025.

Foley & Lardner. “Breaking Down the October 2, 2025 NVCA Updates to the Model Legal Documents.” 2025.

Holland & Knight. “U.S. Companies Face EU AI Act’s Possible Compliance Deadline.” April 2026.

Internal Revenue Code Section 83(b).

Internal Revenue Code Section 1202, as amended by the One Big Beautiful Bill Act, Public Law 119 21. July 4, 2025.

IPWatchdog. “U.S. Patent Litigation Trends in 2025: Patterns Behind the Numbers.” September 28, 2025.

National Institute of Standards and Technology. AI Risk Management Framework Generative AI Profile. NIST AI 600 1. July 26, 2024.

National Venture Capital Association. Model Legal Documents. Updated October 2025.

OpenEvidence, Inc. v. Pathway Medical, Inc. U.S. District Court for the District of Massachusetts, No. 1:25 cv 10471, filed February 2025.

S.G. Borello & Sons, Inc. v. Department of Industrial Relations, 48 Cal. 3d 341 (1989).

State of California, et al. v. Mullin, et al. U.S. District Court for the District of Massachusetts. Decided June 8, 2026.

U.S. Citizenship and Immigration Services. Policy Manual Volume 2, Part M, Policy Alert PA 2025 02. January 8, 2025.

U.S. Department of Homeland Security. Weighted Selection Process for Registrants and Petitioners Seeking to File Cap Subject H 1B Petitions. Federal Register. December 29, 2025.

U.S. Department of Labor. “Employee or Independent Contractor Classification Under the Fair Labor Standards Act.” Federal Register. January 10, 2024.

U.S. Department of Labor. Notice of Proposed Rulemaking on Employee or Independent Contractor Status. Federal Register. February 26, 2026.

U.S. Department of Labor, Wage and Hour Division. Enforcement policy announcement. May 1, 2025.

U.S. Department of State. Visa Bulletin for July 2026.

U.S. Patent and Trademark Office. Inventorship Guidance for AI Assisted Inventions. February 13, 2024, and November 2025 revision.

Interested in analysis about the intersection of corporate law, securities regulation, tech, artificial intelligence, public policy and the law? Check out my Substack channel. https://theinnovationattorney.substack.com/