The AI Revolution: Lawsuits, New Jobs, and Shifting Sentiments

The world of artificial intelligence continues its rapid evolution, marked by high-stakes legal battles, the emergence of critical new job roles, and a growing debate about AI's impact on employment and society. This week, we delve into the ongoing legal drama surrounding Elon Musk and OpenAI, explore the rise of "Forward Deployed Engineers," and examine the increasingly polarized public sentiment towards AI.

Musk v. OpenAI: The Jury Deliberates

The highly publicized lawsuit filed by Elon Musk against OpenAI has reached a critical juncture, with a jury now deliberating closing arguments. The trial has laid bare internal disputes and personal communications between key figures in the AI industry.

Ilia Sutzkver, former chief scientist at OpenAI, testified that he spent a year gathering evidence of Sam Altman's "consistent pattern of lying" and had prepared a detailed document for the OpenAI board outlining his concerns. Sutzkver's OpenAI stock is now reportedly valued at approximately $7 billion. Microsoft CEO Satya Nadella also testified, asserting that Microsoft's $13 billion investment did not violate OpenAI's nonprofit mission and that Musk never raised objections despite having Nadella's contact information.

Sam Altman himself testified that Musk repeatedly attempted to merge OpenAI with Tesla or convert it into a for-profit entity where Musk would hold majority ownership. One particularly "hair-raising" moment, according to Altman, involved Musk suggesting that control of OpenAI could pass to his children upon his death.

Musk's legal team accused Altman and OpenAI President Greg Brockman of "stealing a charity," while OpenAI's lawyers countered that Musk's primary concern was "winning," not the nonprofit structure, presenting evidence of Musk's own proposals for a for-profit OpenAI with over 50% ownership. Musk is seeking around $150 billion in damages, Altman's removal from the board, and the unwinding of OpenAI's for-profit conversion.

While the jury's role is advisory, their recommendations will significantly influence the judge's final decision. The trial has undeniably exposed private communications and internal workings of major AI players, offering a rare glimpse into the personal dynamics and financial stakes involved. However, it remains unclear whether either side has emerged victorious in the court of public opinion.

"Forward Deployed Engineers" Emerge as AI's Hot New Job

The enterprise AI landscape is witnessing the rapid ascent of a new critical role: the "Forward Deployed Engineer" (FDEE). These engineers embed themselves within customer organizations to design and implement AI systems alongside frontline teams.

OpenAI has significantly amplified this trend with the launch of its "deployment company," a new business unit backed by over $4 billion in initial investment. This venture aims to embed FDEEs within customer organizations to identify high-value AI workflows, redesign processes, and establish robust production systems. The acquisition of applied AI consulting firm Tomorrow.io, bringing 150 experienced FDEEs aboard, further solidifies this strategic move.

Simultaneously, Google Cloud announced a new AI-focused organization within its go-to-market team, planning to hire a substantial number of FDEEs to drive customer AI transformation. This initiative is supported by a $750 million ecosystem commitment to aid Google's partner network in deploying agentic AI.

Industry leaders are recognizing the significance of this role. Box CEO Aaron Levy predicts FDEEs will become one of the most in-demand jobs in tech, emphasizing that deploying AI agents is more technically demanding than traditional software deployment, requiring deep understanding of customer business processes. However, AI advocate Ali K. Miller cautions that relying solely on FDEEs for transformation can be costly, highlighting the necessity of broader change management, communication, and education.

The FDEE role is essentially a highly technical consultant capable of customizing AI models and agents to solve specific business problems. This allows for outcome-based pricing, where fees are tied directly to the business value generated. While consulting firms like Deloitte and McKinsey have historically filled this niche, AI companies are now directly competing for this talent, recognizing the immense revenue potential in the $6 trillion knowledge work labor market.

The need for FDEEs stems from the fact that many enterprises lack the internal expertise to effectively adopt and implement advanced AI models and agentic capabilities. This functional need is amplified by the capitalistic drive of AI companies to generate revenue and justify their valuations. As Paul Ritzer notes, this is a logical move for companies needing to scale AI adoption rapidly, even if it means competing directly with traditional consulting partners for talent.

The AI Jobs Apocalypse Debate Intensifies

A chorus of prominent voices has recently pushed back against the narrative of an impending AI-driven "jobs apocalypse." Figures from venture capital, academia, and journalism are arguing that fears of widespread job loss are largely manufactured or exaggerated.

Scott Galloway, in a widely shared essay, framed the AI jobs apocalypse as a "narrative-driven" strategy engineered by those who profit from fear. He pointed to the stability of US tech employment and attributed recent layoffs to a return to pre-pandemic headcount, rather than AI displacement. Galloway suggests the narrative is a marketing tactic to inflate the perceived value of AI products.

David George of Andreessen Horowitz likened the AI doomer position to the "lump of labor fallacy," arguing that historical precedent shows technological advancements create new industries and jobs, rather than simply reducing the total amount of work. Andrew Ng echoed this sentiment, highlighting strong hiring in software engineering, a field directly impacted by AI coding tools, and pointing to a healthy overall US unemployment rate. He also suggested AI labs have an incentive to overstate AI's job-displacing capabilities to justify higher pricing.

Derek Thompson, in a podcast titled "The Smartest Case Against the AI Jobs Apocalypse," argued that human desire and status-seeking are insatiable, leading to the emergence of new work categories even as AI automates existing tasks.

However, Brookings researcher Molly Kinder offers a more nuanced perspective, warning that the focus on a binary between the current labor market and a future post-AGI world of abundance ignores a significant period of disruption. She suggests AI could reverse the skill premium for knowledge workers, with losses concentrated in high-paying cognitive roles.

Our own research at Smarter X paints a starkly different picture. In our latest "State of AI for Business" report, 71% of respondents believe AI will eliminate more jobs than it creates in the next three years, a significant jump from previous years. This sentiment is consistent across all levels of an organization, from CEOs to specialists. This growing concern among professionals, particularly those actively using AI, suggests a disconnect between the optimistic pronouncements of some industry leaders and the lived experience and anxieties of the workforce.

The reality on the ground, as seen with General Motors laying off IT staff to "swap skills" for AI-focused roles, indicates a shift in demand. While GM is replacing 600 IT employees, they plan to hire 200 with AI-specific backgrounds, suggesting a future where AI literacy is paramount, but the overall number of employees needed may decrease. This trend, coupled with the increasing realization by figures like Ken Griffin, CEO of Citadel, that AI is "profoundly more powerful" and capable of automating complex white-collar tasks, suggests that the debate over AI's impact on jobs is far from settled.

An AI Hate Wave Is Here

A palpable backlash against artificial intelligence is emerging in the United States, with data indicating a significant shift in public sentiment. A recent Gallup survey found only 18% of young people feel hopeful about AI, while an Economist/YouGov poll revealed over 70% of Americans believe AI is advancing too quickly. Negative views of AI have risen from 34% three years ago to over 50% today.

This sentiment is manifesting in tangible ways, such as a record number of data center cancellations in the first quarter of 2026 due to community resistance. Analysts at Morgan Stanley note that public pushback is becoming a "binding constraint" on investments, particularly concerning data center buildouts.

The backlash is particularly evident at public events. A commencement address at the University of Central Florida saw a real estate executive met with boos when she described AI as the next industrial revolution. Similarly, former Google CEO Eric Schmidt faced a chorus of boos at the University of Arizona when he drew parallels between AI and the transformative impact of the computer, a comparison that was met with fear and apprehension from the graduating class. Even when Schmidt attempted to highlight AI's positive applications, such as its potential to solve cancer, the boos intensified.

This growing public distrust presents a significant challenge for the AI industry, which has been criticized for downplaying the societal impact of its technology in favor of a "future of abundance" narrative. The widespread negative sentiment could embolden politicians to capitalize on these fears, potentially leading to increased regulation and public scrutiny.

For individuals entering the workforce, this evolving sentiment poses a critical dilemma. While embracing AI is becoming essential for career advancement, the prevailing negative perception could create a challenging environment for those seeking employment in AI-related fields. The fear and anxiety surrounding AI's impact on jobs and society are real, and ignoring them could lead to further societal division and a missed opportunity for responsible AI integration.

Two Scenarios Could Unfold in the US-China AI Race

Anthropic's recent paper, "2028: Two Scenarios for Global AI Leadership," outlines two potential futures for the US-China AI race, emphasizing the urgency for policymakers to act.

In the first scenario, the US defends its compute advantage by tightening chip export controls, disrupting China's efforts to distill US models, and accelerating democratic AI adoption. This pathway, Anthropic projects, could secure a 12 to 24-month US lead in AI by 2028.

The second scenario depicts a failure by the US to act, allowing China's Communist Party to achieve near-frontier intelligence, deploy subsidized AI globally, and shape the rules and norms of the technology under authoritarian regimes.

Anthropic warns that the window to choose a path is rapidly closing, suggesting that 2026 may be remembered as the critical "breakaway opportunity" for American AI. The paper highlights Anthropic's own model, "Mythos," as a wake-up call for the accelerating pace of AI development.

Meanwhile, a summit between President Trump and Chinese leader Xi Jinping yielded no signed AI agreement. Treasury Secretary Scott Bessant indicated discussions would continue on best practices for AI safety, particularly concerning non-state actors accessing models. Trump mentioned discussions about potential collaboration on guardrails. Notably, Nvidia CEO Jensen Huang, initially excluded from Trump's delegation to China due to optics surrounding chip controls, was later invited aboard Air Force One.

The timing of Anthropic's paper release during Trump's visit underscores the geopolitical significance of AI. The potential for a widening US lead or China's ascent to global AI dominance hinges on strategic policy decisions made in the immediate future.

AI Threats Have the US Government (and Labs) Worried

The escalating cybersecurity threats posed by AI have prompted bipartisan concern among US lawmakers. A letter from 32 House members urged the White House to address AI-driven cybersecurity risks, particularly in light of Anthropic's Mythos model identifying thousands of high-severity zero-day vulnerabilities.

Palo Alto Networks, an early access partner for Mythos, reported finding seven times its normal rate of vulnerabilities in its own products in the past month, estimating that attackers could gain broad access to these capabilities within three to five months.

Google's threat intelligence group has documented the first known instance of cybercriminals using AI to develop a zero-day exploit in the wild, warning that the "AI vulnerability race has already begun." Their report also details the use of AI by state-backed hackers from Russia, North Korea, and Beijing to scale up cyberattacks.

In response, OpenAI has launched "Daybreak," a cybersecurity initiative providing cyber-capable versions of GPT-5.5 to partners like Cloudflare, Cisco, Crowdstrike, and Palo Alto Networks. The Trump administration has also been considering executive action on frontier model cybersecurity, though internal disagreements and the China trip have reportedly delayed this process.

The rapid advancement of AI in cybersecurity presents a daunting challenge. As Paul Ritzer notes, if organizations are not forward-deployed engineers, becoming cybersecurity experts might be the next best career path. The speed at which open-source models are likely to adopt these advanced capabilities, combined with the potential for prolific misuse, suggests a race against time to develop effective defenses.

The Rise of "Headless" Software

The increasing autonomy of AI agents is prompting a fundamental reevaluation of traditional software architecture and business models. Salesforce's recent launch of a "headless" product, opening its APIs and betting on its data layer rather than its user interface (UI) as its core value, exemplifies this shift.

Andreessen Horowitz (a16z) argues that as agents become the primary users of business software, the traditional UI's role in enforcing data hygiene, creating shared vocabulary, and building user muscle memory becomes less critical. Agents can read and write directly to underlying data, bypassing the need for a human-centric interface.

a16z suggests that defensibility in this new landscape will shift towards operational logic, compliance, critical data, and connectivity across systems. For AI-native startups, proprietary data, owning the full action loop, network effects, and real-world execution will become paramount. The future of systems of record will likely involve agentic systems that capture context, initiate actions, and generate their own data exhaust, extending into real-world execution and multi-party workflows.

This trend poses a significant challenge for established software companies. If agents can access and utilize data directly through APIs, the need for traditional UIs diminishes, potentially eroding the value proposition of platforms built around human interaction. The uncertainty surrounding this transition is reflected in the stock performance of software companies compared to AI-focused firms, highlighting the ongoing evolution of the software industry in the age of intelligent agents.

Publicis Acquires LiveRamp

In a significant move to position itself for the burgeoning agentic transformation market, Publicis Group has agreed to acquire LiveRamp for $2.2 billion in an all-cash deal. Publicis, a major advertising holding company, views this acquisition as a cornerstone of its strategy to lead in the use of AI agents for automating and collaborating on corporate workflows, a market estimated to be worth approximately $1 trillion.

LiveRamp specializes in data collaboration, enabling companies to securely share and build data sets and models together. Publicis frames these capabilities as a foundational element for powering agentic AI frameworks. This strategic acquisition underscores the growing recognition within the advertising and marketing industry that data infrastructure and AI agent capabilities are intrinsically linked for future growth and innovation.

How AI Is Changing the Way We Work

Several recent developments highlight the profound ways AI is reshaping how we approach our work.

Kieran Flanigan, a go-to-market leader, has detailed his creation of an "AI second brain" using Obsidian and Claude code. This system loads his strategic context and notes at the start of each session, allowing him to query and receive structured summaries of his work and information, effectively creating a conversational database of his professional knowledge.

Shopify CEO Toby Lütke shared how his company built an AI agent named "River" that operates in public Slack channels. River has engaged with thousands of employees and authored a significant portion of merged pull requests into Shopify's codebase, demonstrating how AI can be integrated into collaborative workflows and provide transparency across an organization.

AI researcher Andre Karpathy shared a practical tip: asking LLMs to structure their responses in HTML rather than markdown. Karpathy argues that HTML offers a more visually intuitive way for the human brain to process information, enhancing the review and understanding of AI-generated output.

These examples, from personal AI assistants to organizational agents and improved output formats, illustrate a broader trend: AI is not just automating tasks but fundamentally altering how professionals learn, contribute, and advance. The traditional model of professional development, where junior employees learned through foundational tasks now being absorbed by AI, is being challenged. This necessitates a rethinking of organizational structures and talent development, potentially giving rise to new roles focused on designing, orchestrating, and learning within AI-augmented systems.

AI Product and Funding Updates

The AI landscape continues to be dynamic with significant product launches and funding rounds:

Key Takeaways