Tech Investor Backs Risky Startups While Spending Millions to Shape Federal Artificial Intelligence Policy
Menlo Park, Tuesday, 25 August 2026.
Venture firm Andreessen Horowitz is spending heavily to loosen federal AI regulations, despite funding high-risk startups that operate illegal social media bot farms and deceptive platforms.
A Dual Strategy of Lobbying and Government Influence
The political maneuvering of Andreessen Horowitz (a16z) has reached unprecedented levels as the firm seeks to shape the future of artificial intelligence governance in Washington. In August 2025, the firm launched “Leading The Future,” a super PAC designed to align with federal AI policy objectives [1]. By April 1, 2026, the super PAC had raised $125 million to influence the 2026 midterm elections, drawing financial support from a16z, Greg Brockman, Joe Lonsdale, and Perplexity [2]. This lobbying push is accompanied by a strategic placement of former partners into crucial government roles. Sriram Krishnan, an a16z general partner until shortly before his appointment, became the Senior White House Policy Advisor on AI in December 2024, working alongside David Sacks, the White House AI and crypto czar [1]. Additionally, in August 2025, former a16z partners Scott Kupor and Jamie Sullivan assumed influential roles within the Office of Personnel Management and the Department of Government Efficiency, respectively [1].
Federal Preemption and the Push for Deregulation
This concentrated influence in Washington has yielded substantial regulatory victories for the venture capital giant. In December 2025, President Trump signed an executive order intended to undermine state-level AI regulations, effectively preempting state laws in favor of a more permissive federal framework [1]. This policy shift aligns directly with a16z co-founder Marc Andreessen’s public stance; in his October 2023 “Techno-Optimist Manifesto,” Andreessen characterized risk management, tech ethics, and the precautionary principle as “enemies” of progress, even going so far as to claim that slowing down AI development is a “form of murder” [1]. However, this aggressive push for deregulation stands in stark contrast to public sentiment. A Pew Research poll from April 3, 2025, revealed that 58% of Americans believe government regulation of AI is insufficient, while only 21% believe it goes too far [1].
Funding the Underbelly of Social Media Manipulation
While a16z campaigns for a deregulated AI future in Washington, its investment portfolio contains startups that actively exploit regulatory loopholes and engage in deceptive practices. In October 2025, a16z invested $1 million through its Speedrun program in Doublespeed, a startup that operates a “phone farm” of over 1,100 devices and 400 TikTok accounts [1]. The company utilizes AI-driven bot networks to mimic human behavior on social media, bypassing platform restrictions to promote products [1]. Doublespeed’s founder, Zuhair Lakhani, openly celebrated this model, stating, “We run the only VC-backed bot farm in America. Because why let Russia and China have all the fun?” [1]. A December 2025 data breach of Doublespeed’s backend exposed that its AI-generated influencers were actively promoting herbal supplements and wellness products to older demographics by feigning medical conditions, violating both TikTok Community Guidelines and Federal Trade Commission (FTC) regulations [1].
The Ethics of Deception and Academic Dishonesty
Doublespeed is not the only portfolio company raising ethical alarms. In June 2025, a16z led a $15 million Series A investment in Cluely, a startup whose original marketing manifesto openly advocated for “cheating” on sales calls, meetings, and negotiations [1]. Cluely’s co-founder and CEO, Roy Lee, had previously developed “Interview Coder,” an AI tool used to cheat on technical hiring interviews, which resulted in him being placed on academic probation by Columbia University in early 2025 [1]. Facing intense public scrutiny, Cluely scrubbed references to academic and professional dishonesty from its website by November 2025, repositioning itself as an “AI meeting assistant” [1]. However, the startup continues to advertise its services as “undetectable,” and Lee has openly embraced controversial marketing, stating that saying “extremely crazy shit online” drives downloads and eyeballs [1].
Exploiting Loopholes in Gambling and Synthetic Media
The firm’s appetite for high-risk, legally ambiguous models extends into financial speculation and synthetic media. In October 2025, a16z co-led a $300 million Series D for Kalshi, followed by a $1 billion Series E in December 2025 that valued the platform at $11 billion [1]. Kalshi bypassed federal and state gaming restrictions by offering “futures contracts” rather than traditional sports betting, allowing 18-year-olds to wager in jurisdictions where the legal gambling age is 21 [1]. As of August 2026, Kalshi operates across all 50 states, despite a November 2025 Nevada federal court ruling that rejected its legal interpretations as “strained” [1]. Other a16z-backed gambling ventures have faced immediate regulatory crackdowns; for instance, Arizona’s Department of Gaming issued a cease-and-desist order to BettySweeps in August 2025, labeling it a “felony criminal enterprise” [1], while California Governor Gavin Newsom signed legislation in September 2025 banning sweepstakes platforms like Cheddr [1].
Severe Safety and Compliance Failures
Perhaps the most disturbing aspect of the firm’s portfolio lies in its investments in synthetic media and decentralized tech. In June 2023, a16z led a $5.1 million seed round for Civitai, a platform hosting over 35,000 deepfake models, 96% of which depict identifiable women [1]. Civitai has acknowledged 178 reports of AI-generated child sexual abuse material (CSAM) to the National Center for Missing & Exploited Children (NCMEC) and over 252,000 user attempts to bypass restrictions in a single quarter [1]. These ethical failures mirror broader systemic issues within a16z’s fintech investments. The firm’s $33 million Series B investment in Synapse in June 2019 collapsed into bankruptcy on April 22, 2024, leaving up to $96 million in customer funds missing [1]. More recently, in December 2025, a16z led a $34 million Series A in Truemed, a company that automates medical letters so consumers can use tax-advantaged health accounts to purchase luxury items like saunas, a practice the IRS warned in March 2024 does not qualify as tax-deductible [1].