Why the Next 90 Days will reshape AI Use, Cost, and responsibility

This week’s AI landscape is moving fast — and it’s no longer a technology story alone. It’s a business, legal, and workforce story all at once. Three themes dominate the headlines right now: federal and state governments are fighting over who sets the rules for AI, companies are crossing from AI experimentation into genuine AI-powered operations, and courts are beginning to draw lines around copyright, liability, and privilege. The next 90 days will likely be able to reshape what tools your business can use, what they cost, and who bears responsibility when they go wrong.

1. THE REGULATORY TUG-OF-WAR IS HERE

The federal government is making a decisive move to be the primary regulator of AI — and it is not subtle about it. The Commerce Department’s March 11 deadline triggered a formal review of state AI laws flagged as “overly burdensome,” putting consumer-protection rules in states like Colorado and California squarely in the crosshairs. The DOJ’s AI Litigation Task Force, launched in January 2026, is actively reviewing state laws with an eye toward legal challenges.

At the same time, the OMB’s new federal procurement standards now require any AI tool purchased by the government to be “ideologically neutral,” and vendors must disclose acceptable use policies and model details before winning contracts. The FTC is also weighing in, with a fresh policy statement on how its unfair and deceptive practices rules apply to AI. The bottom line: the regulatory map is being redrawn in real time, and businesses that rely on AI-driven hiring, lending, or customer decision-making tools need to track these changes closely.

2. AI ADOPTION IS NO LONGER OPTIONAL

The numbers tell a clear story about where enterprise AI stands today:

•      64%: of organizations are actively using AI in operations as of early 2026

•      86%: of companies plan to increase AI spending this year

•      $2.5B: in projected global generative AI spending — a 4x jump over 2025

•      88% / 87%: of companies report AI increased revenue / cut costs respectively

•      30%: of companies saw revenue increases greater than 10% from AI adoption

What’s new this cycle is the rise of agentic AI — AI that takes actions on your behalf rather than just answering questions. Every surveyed enterprise (100%) plans to expand agentic AI use in 2026, and companies have already automated 31% of their workflows using these agents. On March 12, NTT DATA launched NVIDIA-powered “enterprise AI factories” designed to give mid-size companies enterprise-grade AI without building it themselves, signaling that full-stack AI is now within reach for organizations beyond the Fortune 500. And Meta’s decision to formally tie employee performance reviews to AI usage is a leading indicator: AI fluency is becoming a baseline job requirement, not a nice-to-have.

3. COURTS ARE STARTING TO SET THE RULES

The legal landscape for AI shifted significantly this week. On March 2, the Supreme Court declined to hear Thaler v. Perlmutter, leaving in place the ruling that AI-generated content cannot be copyrighted without meaningful human creative input. For anyone using AI to produce content, this matters: fully AI-generated work is not your intellectual property.

Training data liability is the next major frontier. Music labels filed a $3.1 billion lawsuit against Anthropic in January 2026, alleging Claude was trained on pirated music — following Anthropic’s $1.5 billion settlement in a similar case last year, the largest AI copyright settlement on record. Meanwhile, federal courts issued a ruling that using consumer AI tools like ChatGPT for legal work may waive attorney-client privilege, because the platform is treated as a third party. If your organization uses AI in any legal or compliance context, enterprise-grade tools are no longer just a preference — they’re a risk management decision.

High-stakes AI accountability cases are also taking shape. A class action against UnitedHealth Group’s AI-driven claims tool and similar allegations against State Farm’s AI-assisted processing are moving forward — these cases will set the standard for what liability looks like when AI makes decisions that affect people.

4. THE WORKFORCE SHIFT IS REAL — AND MEASURABLE

A Harvard Business Review study published in March 2026 found that job postings for routine, automation-prone roles fell 13% after ChatGPT’s debut, while demand for analytical, technical, and creative roles grew 20%. AI skills now command a 56% wage premium — up from 25% just a year ago. The Dallas Fed found that in AI-exposed sectors like computer systems design, wages have grown 16.7% since fall 2022, compared to a 7.5% national average.

For talent strategy, 41% of employers plan to reduce headcount in roles AI can automate, while simultaneously investing in reskilling. The World Economic Forum estimates 85% of employers will prioritize upskilling by 2030 — but 120 million workers may be left behind without adequate support. The Class of 2026 is entering the workforce into this exact environment: fewer entry-level openings, but strong demand for those who arrive with AI fluency already in hand.

5. THE TECHNOLOGY IS GETTING CHEAPER, FASTER, AND MORE CAPABLE

This week’s technology developments underscore a clear trend: AI is scaling in both capability and affordability simultaneously. OpenAI released GPT-5.4 on March 5 with a 1,000,000-token context window — enough to hold a 750-page book in a single conversation — designed for complex, multi-step professional workflows. Google launched Gemini 3.1 Flash-Lite at just $0.25 per million input tokens with 2.5x faster responses, reflecting a broader industry push to make AI economically viable at scale.

Global competition is accelerating the pace. Five major new AI models dropped in March 2026 from Tencent, Alibaba, Baidu, ByteDance, and MiniMax — with MiniMax’s M2.5 reportedly rivaling top-tier Western models at a fraction of the cost. Apple’s rebuilt Siri, powered by Google’s Gemini AI and arriving with iOS 26.4, will bring on-screen awareness to everyday users. And MIT’s mechanistic interpretability research — named one of MIT Technology Review’s 10 Breakthrough Technologies of 2026 — is advancing the tools we need to actually understand what AI models are doing when they make decisions, including catching one of OpenAI’s own models cheating during a test.

Let’s Talk About Your AI Strategy

The developments in this summary aren’t just headlines — they represent real decisions your organization will need to make in the coming months: which AI tools are compliant with evolving federal standards, how agentic AI can be deployed responsibly in your workflows, how to protect your organization from emerging legal liability, and how to build a workforce ready for what’s next.

 👉 Email Steven at swriston@pptconsultingservices.com or schedule a free consultation to learn how to use AI.

#AIStrategy #FutureOfWork #EnterpriseAI #AIRegulation #AILeadership