Desiree Sainthrope, a legal powerhouse with deep roots in international trade and global compliance, joins us to dissect the increasingly complex relationship between tech giants and the regulatory frameworks that govern them. In 2026, the battle for the soul of artificial intelligence is no longer confined to the lab; it is being fought in the halls of power through strategic lobbying, safety narratives, and significant capital shifts. As an authority on the evolving implications of technology policy, Desiree brings her unique perspective to recent warnings regarding the concentration of market power and the growing intersection of autonomous systems and modern defense. We explore the tension between ensuring technological safety and preventing the rise of an industry oligopoly, the strategic maneuvers of top-tier investors, and the ethical weight of human accountability in an era of automated warfare.
Major AI developers often fund outside groups to advocate for stricter industry rules; how do you see this strategy impacting the delicate balance between public safety and fair market competition?
This strategy is a classic example of what we in legal compliance call “regulatory capture,” where the very entities meant to be regulated end up shaping the laws to suit their own survival. When organizations like OpenAI or Anthropic push for rigid oversight, it creates a high-stakes environment where safety concerns—though very real—are used as a lever to centralize control over public policy. By funding external advocacy groups, these firms can effectively build a defensive wall of compliance that smaller, more nimble startups simply cannot afford to climb, thereby entrenching the market power of the dominant players. We have to be incredibly careful because while the “fear and regulation” narrative is compelling to the public, it risks creating an oligopoly that dictates the terms of innovation to the government. The danger lies in a handful of firms holding the keys to the framework, potentially stifling the competitive diversity that is essential for a healthy technological ecosystem.
Given that some investors are simultaneously praising the commercial performance of these companies while calling their regulatory pushes “very dangerous,” how should we interpret this paradox in the current business climate?
It is a fascinating contradiction to see an investor like Joe Lonsdale maintain such a vocal pride in a company like Anthropic while simultaneously attacking its political maneuvering. This suggests a clear separation between a company’s talent and its commercial success—which has been described as some of the highest-performing in the world over the last few years—and the broader strategic choices made by its leadership regarding government relations. From a business journalism perspective, this highlights that you can believe in a product’s brilliance while fearing the long-term impact of its corporate policy on the wider market. It is an acknowledgment that while the company is winning the commercial race, its attempts to influence the rules of the game could have unintended, negative consequences for the industry’s future. This duality reminds us that high-performance metrics do not automatically grant a company’s regulatory agenda a pass on ethical or competitive grounds.
With the deepening partnerships between Western defense firms and international businesses in conflict zones, what are the primary legal and strategic risks when these relationships lead to direct acquisitions?
The trend of firms like Anduril and those backed by 8VC partnering closely with overseas entities, particularly in active regions like Ukraine, marks a significant shift in how defense technology is developed and tested. These relationships are often born out of immediate necessity and provide a “battlefield laboratory” for innovation, but the transition from partnership to acquisition carries immense complexity regarding intellectual property and international trade compliance. When a major firm moves to buy these adapted technologies, they aren’t just buying code; they are buying systems that have been forged in what Lonsdale calls the “fog of war,” where the circumstances of their development are often chaotic. Legally, we have to ensure that these acquisitions don’t bypass rigorous assessment standards or create a feedback loop where military systems are deployed without a clear understanding of their long-term operational risks. The investor interest here is palpable, but the challenge remains in integrating these “field-tested” innovations into a stable, accountable corporate and legal framework.
How can we effectively maintain the principle of human accountability as autonomous weapons become more integrated into military strategy, especially when critics are often told they cannot judge from the outside?
Maintaining human accountability requires a rigid legal structure where every autonomous system is tied to a specific individual who is responsible for its study, work, and ultimate approval. The “fog of war” is frequently cited as a reason to avoid outside judgment, but from a global compliance perspective, we cannot allow the complexity of the battlefield to become a shield against ethical oversight. There must always be a person in charge who can be held liable for the decisions made by an algorithm, ensuring that the technology remains a tool of human intent rather than an independent actor. This means we need more than just technical safety; we need a transparent trail of authorization that survives the chaos of conflict and can be audited after the fact. It is easy to say that outsiders don’t understand the circumstances, but that is precisely why the internal accountability mechanisms must be so robust that they can withstand any level of external scrutiny.
What is your forecast for the future of AI regulation and the role of tech leaders in government?
I anticipate that the next two years will see a massive pushback against the current trend of concentrated policy influence, as smaller tech firms and civil society demand a more democratic approach to rule-making. We will likely see a formalization of “human-in-the-loop” mandates for both military and high-stakes civilian AI, moving away from vague safety promises toward concrete, legally binding accountability standards. Regarding the role of tech leaders in the Pentagon or other senior government positions, I believe we are entering an era of the “citizen-builder,” where figures like Lonsdale may be called upon by a President to serve, but will face unprecedented scrutiny over their previous business ties and investment portfolios. Ultimately, the industry will have to choose between the short-term gains of an oligopoly and the long-term stability of a transparent, competitive market that operates under clear, human-centered laws. This tension will define the regulatory landscape for the remainder of the decade.
