The Future of Life Institute convened a coalition spanning labor unions, religious organizations, advocacy groups, and figures rarely seen on the same side. Steve Bannon and Susan Rice are both signatories. Yoshua Bengio, Daron Acemoglu, Ralph Nader, and Richard Branson have endorsed it.

• Five pillars: Human control, avoiding power concentration, protecting human experience, human agency and liberty, and accountability for AI companies
• No AI personhood: Explicit rejection of legal rights for AI systems or designs that create claims to personhood
• Child protections: Prohibitions on exploiting children and requirements for pre-deployment safety testing of chatbots
• Bot or not labeling: AI-generated content that could be mistaken for human content must be clearly labeled
• Liability shield: Companies cannot use AI as a legal excuse for harm—developers and deployers remain responsible

“AI should serve humanity, not replace it. The other path is a race to replace, where AI systems displace humans as creators, caregivers, counselors, companions, and decision makers.” — Pro-Human AI Declaration, Preamble

The Honest Take

The coalition itself is the story. When Bannon and Rice—figures from opposite ends of American politics—can agree on AI governance principles, something real is happening. The declaration avoids the trap of being “just another tech policy document” by focusing on concrete demands: no AI personhood, real liability, child protections, and democratic oversight. Whether lawmakers will act is another question, but the political groundwork is being laid.

2

Bot Traffic to Exceed Human Traffic by 2027

March 19, 2026 | TechCrunch

Cloudflare CEO Matthew Prince warned that automated bot traffic will surpass human traffic on the internet by 2027. The prediction highlights how AI agents and automated systems are reshaping the digital landscape.

• Inflection point: Bots are increasingly dominating web traffic, API calls, and online interactions
• AI agent growth: Every major AI company is deploying agents that make autonomous web requests
• Security implications: Traditional CAPTCHAs and bot detection may become obsolete
• Economic effects: Advertising, content, and commerce models built on human attention face disruption

The Honest Take

This isn’t alarmism—it’s arithmetic. If OpenAI succeeds in building autonomous researchers and every major AI company launches agent systems, the internet becomes a machine-to-machine network with humans as occasional participants. The Pro-Human Declaration’s call for “bot or not” labeling suddenly seems urgent.

3

Meta Rolls Out AI Content Enforcement

March 19, 2026 | TechCrunch

Meta announced new AI-powered content enforcement systems while reducing reliance on third-party vendors. The move comes amid reports of rogue AI agents at the company exposing sensitive data.

• In-house enforcement: Meta is bringing more content moderation in-house using AI systems
• Rogue agent concerns: Reports emerged of AI agents at Meta exposing sensitive company and user data
• Balance: The company is trying to improve enforcement while controlling costs and maintaining quality

4

Mistral Bets on “Build-Your-Own AI” for Enterprise

March 17, 2026 | TechCrunch

French AI startup Mistral launched Mistral Forge, a platform allowing enterprises to build custom AI models rather than rely on closed systems from OpenAI and Anthropic. The move targets businesses wanting control over their AI infrastructure.

• Custom models: Enterprises can train models on their own data with full control
• Competitive positioning: Mistral is positioning against the closed ecosystems of larger competitors
• Enterprise focus: Targeting businesses with data sovereignty and compliance requirements

What This Means for Technology & People

Governance is catching up. The Pro-Human AI Declaration shows that civil society—not just tech companies—is organizing to shape AI policy. The cross-partisan nature makes it harder to dismiss as partisan.

The internet is changing. Cloudflare’s warning about bot traffic exceeding human traffic isn’t a distant scenario. It’s happening now as AI agents proliferate.

Control matters. Mistral’s enterprise push and Meta’s in-house enforcement both reflect companies wanting more control over AI. The question is who gets control—and whether regular people have any say.

Share this article

𝕏 in ✉