Three developments this week are reshaping the AI and startup legal landscape in ways that will outlast any single news cycle. The Trump administration is on the verge of signing an executive order that would formalize government oversight of frontier AI models — but an internal battle between intelligence agencies and the Commerce Department is actively reshaping the order's scope and who controls it. Iowa became the first Midwestern state to enact a chatbot safety law, joining Connecticut and Utah in a growing wave of state AI content legislation. And the INVEST Act — the most significant capital formation reform in over a decade — is now in the Senate, carrying provisions that could fundamentally change how startups raise early-stage capital. None of these are future developments. All three require founders to act or plan now.

The Big Picture

The most consequential AI regulatory development of 2026 may arrive this week without a scheduled announcement. The Trump administration's AI executive order — which has been in draft form for weeks — is expected to be signed imminently, possibly as early as Monday. What makes this week's news distinct from prior reporting is not the existence of the order: it is what the internal battle over the order reveals about who will actually control AI oversight in the United States. The answer to that question will determine the compliance burden for every AI company in America for the next several years.

1. AI and Emerging Tech — The Trump AI Executive Order: Intelligence Agencies vs. Commerce in a "Knife Fight"

Reporting published May 11-12, 2026 by The Washington Post and The Detroit News revealed the full scope of the internal White House battle over the imminent AI executive order. The fight is not over whether to regulate frontier AI — that question has been settled. The fight is over which part of the federal government will have authority to evaluate AI models before they are released to the public.

The two sides of the internal battle:

1. The Commerce Department position — the Commerce Department's AI Safety Institute (now rebranded as CAISI — the Center for AI Standards and Innovation) has spent the past 18 months building an AI evaluation infrastructure, hiring AI experts with advanced degrees, and developing testing frameworks for frontier model capabilities; Commerce argues it is the appropriate home for AI oversight because it has technical expertise and a mandate to support U.S. economic competitiveness
2. The intelligence agency position — the White House's Office of the National Cyber Director, backed by the Office of the Director of National Intelligence (ODNI), is pushing for a large new center within ODNI that would evaluate frontier AI models before public release; intelligence agencies argue they need pre-release access to models like Anthropic's Mythos — which demonstrated the ability to identify and exploit cybersecurity vulnerabilities faster than human hackers — to assess national security risks before those capabilities reach bad actors
3. The trigger: Anthropic's Mythos model — the catalyst for the executive order's acceleration is Anthropic's Mythos, which CISA and national security officials have described as capable of enabling even inexperienced hackers to launch complex attacks on energy grids, banks, and government agencies; this capability profile made frontier AI model oversight a national security imperative, not just a consumer protection concern
4. Bloomberg's reporting on the order's current draft — as of May 8, Bloomberg reported that internal White House drafts of the executive order are trending toward omitting mandatory pre-release testing requirements, replacing them with voluntary government evaluation agreements; this is a materially lighter-touch framework than the FDA-style mandatory review that was initially reported
5. What the order is expected to do regardless of which side wins — the executive order will formalize government evaluation of frontier AI models in some form; it will create a mechanism for pre-release government access to model capabilities; and it will define the category of AI considered a national security concern — a definition that will migrate into future legislation

The question of whether ODNI or Commerce controls AI evaluation is not procedural. It is the difference between a national security-dominated regulatory regime — opaque, intelligence-driven, focused on restricting capabilities — and a technology-competitiveness-dominated regime — transparent, standards-based, focused on enabling innovation with guardrails. For AI founders, the outcome of this internal fight will determine whether their compliance obligations are primarily commercial (working with NIST standards and Commerce frameworks) or national security-adjacent (navigating intelligence community evaluation processes).

Why the AI Executive Order Matters for Founders — Even Those Not Building Frontier Models

Most AI startups are not training frontier models. They are building applications on top of existing foundation models via API. But the executive order's impact on application-layer founders is significant for three reasons. First, the "high-risk AI" definitions the order uses will almost certainly be adopted by Congress in subsequent AI legislation — the categories of AI that require pre-deployment review at the federal level will become the categories regulators focus on at every level. Second, if frontier model providers face pre-release government review delays, the API capabilities available to application-layer startups arrive on a slower timeline than today's rapid release cadence. Third, any executive order that defines national-security-relevant AI capabilities creates a category that enterprise customers — particularly government contractors and regulated financial institutions — will require their AI vendors to certify against. That certification requirement will flow down through enterprise sales and vendor agreements to startups that supply AI tools to covered enterprises.

2. AI and Emerging Tech — Iowa's Chatbot Safety Law: The Midwest Joins the State AI Wave

On May 3, 2026, Iowa Governor Kim Reynolds signed Senate File 2417 — the Iowa Chatbot Safety Law — making Iowa the first Midwestern state to enact a comprehensive framework governing public conversational AI services. Both chambers of the Iowa legislature passed the bill unanimously. The law takes effect July 1, 2027, with enforcement authority vested in the Iowa Attorney General.

What Iowa's Chatbot Safety Law requires:

1. Mandatory AI disclosure — operators of covered conversational AI services must clearly disclose to users that they are interacting with an AI, not a human; this applies to all covered public-facing chatbot products
2. Minor-specific protections — operators must implement privacy management tools specifically for minor users and their guardians, and must establish protocols for interactions involving self-harm, suicide, or mental health crises
3. Prohibition on manipulative engagement tactics — the law bars operators from using AI design techniques specifically engineered to create emotional dependence, extend engagement beyond the user's stated intent, or exploit psychological vulnerabilities
4. Restriction on mental health service impersonation — operators are expressly prohibited from presenting or allowing their AI to present itself as a licensed mental health provider
5. Content restrictions for minors — covered operators must restrict inappropriate content access for minor users and implement age-appropriate content controls

Iowa's law follows Connecticut's SB 5 (companion chatbot provisions effective January 1, 2027), Utah's 2026 AI legislative package, and the federal Take It Down Act (enforcement begins May 19). With Iowa's signature, the state AI content law patchwork now covers the Southeast, Mountain West, Northeast, and Midwest. Founders building chatbot products — particularly companion AI, mental health-adjacent applications, or any conversational AI with significant minor user demographics — now face a multi-state compliance obligation that is effectively a national standard in everything but name.

3. Corporate and Securities Law — The INVEST Act in the Senate & SEC AI Examination Focus

The INVEST Act — the Incentivizing New Ventures and Economic Strength Through Capital Formation Act — passed the House with a 302-123 bipartisan vote earlier this year and is now in the Senate. The bill represents the most significant overhaul of U.S. capital formation rules since the JOBS Act of 2012, with provisions that directly affect how early-stage founders raise capital and how venture funds are structured.

INVEST Act provisions most relevant to founders and emerging funds:

1. Expanded Regulation Crowdfunding — the bill raises the accountant review threshold for Regulation Crowdfunding offerings from $100,000 to $250,000 (with discretion up to $400,000), reducing the audit burden for early-stage companies using crowdfunding as a primary capital-raising mechanism
2. Broader accredited investor definition — the bill expands who qualifies as an accredited investor beyond pure net worth and income tests, adding financial sophistication and professional credential pathways; this expands the pool of eligible investors for Reg D private placements
3. Qualifying venture capital fund expansion — the qualifying venture capital fund size increases from $10 million to $50 million while the investor cap rises from 250 to 500; this creates meaningful new flexibility for emerging managers and micro-VC funds that previously faced registration requirements at modest AUM levels
4. Investment Adviser Act threshold increase — the SEC registration exemption threshold rises from $150 million to $175 million with inflation indexing, reducing the regulatory burden for smaller RIAs and emerging fund managers
5. Regulation A+ modernization — the bill updates Regulation A+ to make it a more viable primary capital-raising tool for growth-stage companies that are too large for crowdfunding but not yet ready for a full IPO

Separately, the SEC's 2026 examination priorities for registered investment advisers and broker-dealers include a specific focus on registrants' use of automated investment tools and AI technologies. Firms using AI in portfolio management, client communications, compliance monitoring, or trading are on the SEC examination radar — and examiners have been instructed to assess whether firms' AI use disclosures, conflicts of interest management, and supervisory frameworks keep pace with their AI adoption. For VC funds and RIAs using AI tools in any part of their operations, this examination focus is a compliance obligation to address before an exam, not during one.

4. How Launch Legal Helps Founders Navigate These Developments

For AI founders building chatbot, companion AI, or mental health-adjacent products: the Iowa law's July 1, 2027 effective date gives you 14 months to build compliance. We help founders map their product's interaction design against Iowa's manipulative engagement prohibition, minor-protection requirements, and mental health impersonation restriction — and build compliance that satisfies Iowa, Connecticut, and Utah simultaneously, rather than engineering separate compliance programs for each state's requirements.

For AI product companies selling to enterprise customers: the Trump AI executive order — in whatever form it takes — will create certification requirements that flow into enterprise procurement and vendor agreements. We help AI founders anticipate what capability disclosures, safety documentation, and government evaluation cooperation their enterprise customers will require, and build that documentation infrastructure before the first enterprise procurement officer asks for it.

For early-stage founders planning their next capital raise: the INVEST Act's expanded Regulation Crowdfunding limits and broader accredited investor definition create structuring opportunities that are worth analyzing now, before the bill passes and the new rules take effect. We help founders evaluate whether the INVEST Act's provisions change their optimal raise structure — and whether a Regulation Crowdfunding round, an expanded Reg D pool, or a Regulation A+ offering makes more sense under the new framework than under current law.

For emerging fund managers and micro-VC funds: the INVEST Act's expansion of the qualifying venture capital fund definition — from $10M to $50M, with an investor cap increase from 250 to 500 — is the most significant structural change for emerging managers in over a decade. We help fund managers evaluate whether the new exemption thresholds change their fund structure, LP agreement terms, or SEC registration obligations, and plan fund formation to take advantage of the expanded exemptions when the bill passes.

What Founders Should Think About Now

• Frontier AI developers: The executive order could arrive any day — if your model development could be characterized as producing capabilities relevant to cybersecurity exploitation or national security risk, engage AI policy counsel before the order is signed; post-signature is too late to shape how your compliance obligations are framed

• Application-layer AI founders: Review the "high-risk AI" definitions when the executive order is published — those definitions will be adopted by Congress and state legislators; understanding where your product sits within them now is essential for enterprise sales, procurement compliance, and fundraising narratives

• Chatbot and conversational AI founders: Iowa's July 1, 2027 effective date combined with Connecticut's January 1, 2027 date means you have 7-14 months to build multi-state chatbot compliance; the founders who design for compliance now rather than retrofitting later will have materially lower implementation costs

• Founders planning capital raises: Monitor the INVEST Act's Senate committee assignment and markup schedule — if it passes before the August recess, the new Reg Crowdfunding thresholds and accredited investor definitions could affect your Q3 or Q4 raise structure

• VC funds and RIAs using AI tools: The SEC's 2026 examination priority on AI use by registered firms is active — if your fund uses AI in any investment, compliance, or client-facing function and has not documented its AI use policies, conflicts management, and supervisory framework, that gap is an examination finding waiting to happen

• All founders with enterprise customers: Build AI capability documentation and safety disclosure infrastructure now — the Trump executive order will create downstream enterprise procurement requirements that your customers will push to their vendors; the founders who have that documentation ready will close enterprise deals faster than those who have to build it in response to an RFP

Strategic Takeaway

Opportunity → The INVEST Act represents the best capital formation environment for early-stage founders since 2012. If the Senate passes it before the August recess, founders who have done the legal analysis in advance — understanding how the expanded Reg Crowdfunding limits, new accredited investor pathways, and Reg A+ modernization apply to their specific situation — will be able to move to market immediately when the new rules take effect. The founders who wait to do that analysis until after the bill passes will lose weeks to legal review while better-prepared competitors are already raising.

Risk → The internal White House battle over AI oversight authority is the highest-stakes AI policy question of the year, and it is being resolved in real time — possibly this week. If intelligence agencies win the turf fight and ODNI becomes the center of AI model evaluation, the compliance environment for AI companies becomes significantly more opaque, more security-clearance-dependent, and less accessible to startup founders than the Commerce-led alternative. This is a risk that founders cannot mitigate after the executive order is signed — but they can engage with through their industry associations, policy coalitions, and public comment processes while the order is still being shaped.

What Comes Next

Watch for the Trump AI executive order — it could be signed as early as this week, and the text will determine whether Commerce or intelligence agencies control frontier AI oversight. Watch for Iowa's Attorney General to issue implementing guidance on the Chatbot Safety Law ahead of its July 1, 2027 effective date — that guidance will define how the manipulative engagement prohibition and minor-protection requirements are interpreted in practice. Watch for the INVEST Act's Senate committee assignment — the bill's path to the Senate floor runs through the Banking or Commerce committee, and assignment will signal the timeline. And watch for the first SEC examination findings related to AI use by RIAs and broker-dealers — those findings will be the clearest signal yet of what AI governance documentation the SEC considers adequate.

Bottom Line

May 13, 2026 is a week where the most important regulatory developments are happening in real time, not on a published calendar. The Trump AI executive order could arrive before this post is 48 hours old. The CLARITY Act markup is tomorrow. Iowa's chatbot law is already signed. The SEC is already examining AI use by registered firms. Founders who treat this as a week to monitor rather than a week to act are misreading the moment. The regulatory frameworks being built this week will govern AI products, startup capital raises, and fund structures for the next decade. The time to engage is now.

Learn More

At Launch Legal, we advise AI founders, early-stage startups, and emerging fund managers on exactly these kinds of converging regulatory timelines — from chatbot compliance across multiple state laws to INVEST Act fundraising strategy to AI executive order impact assessment. If this week's developments raised questions about your product, your raise, or your fund structure, reach out for a consultation.

Sources & Further Reading:

• Washington Post — In Trump Administration Battle Over AI, U.S. Spy Agencies Seek More Power

• Detroit News — In Turf Battle Over AI, U.S. Spy Agencies Vie for More Sway Than Commerce

• Bloomberg — US Prepares AI Security Order That Omits Mandatory Model Tests

• AI Certs — Iowa Enacts Chatbot Safety Law Protecting Users

• Harvard Law School Forum on Corporate Governance — House Passes Bipartisan Capital Formation Package: The INVEST Act

• Troutman Pepper — Proposed State Privacy and AI Law Update: May 11, 2026