• Overheard Capital
  • Posts
  • EU AI Act Takes Off — What It Means for Founders (And Why Meta Isn’t Happy)

EU AI Act Takes Off — What It Means for Founders (And Why Meta Isn’t Happy)

July 22, 2025

In partnership with

Europe’s ambitious AI regulation is no longer a draft—it’s real. With the EU’s compliance deadline for general-purpose AI models looming in just days, the ripple effects are global.

⚖️ What Comes into Effect

Starting August 2, 2025, companies deploying general-purpose AI (think ChatGPT, Midjourney) in Europe must:

  • Assess and manage systemic risks.

  • Submit technical documentation and training data summaries.

  • Adopt transparency and cybersecurity measures.

  • Adhere to content copyright rules.

That’s not all—high-risk AI systems in domains like hiring or health have tighter rules beginning August 2026, with potentially severe penalties up to 7 % of global revenue.

🔍 Meta and Others Push Back

The compliance code released on July 10 offers voluntary guidelines, but not everyone is on board:

  • Meta publicly rejected the guidance, calling it overly broad and litigious.

  • Over 40 major European firms, including Airbus and BNP Paribas, requested a two-year delay.

  • In contrast, OpenAI, Microsoft, and Google are signaling alignment.

The standoff suggests a growing tug-of-war: EU regulators favor oversight; some US tech firms fear a dampening effect on innovation.

🧩 Why Founders Should Care

  1. Global Reach = Global Responsibility
    Selling AI into Europe means compliance, not optional. Even non-EU startups may find the cost too high to ignore.

  2. SME-Friendly Measures Exist
    The Act includes sandboxes, streamlined documentation, and fee waivers designed to support startups, but they have timelines.

  3. Enforcement Is Coming
    After August, companies will be expected to prove transparency, risk assessments, and incident reporting. No proof = fines.

  4. Competitive Differentiator
    Early compliance can become a trust signal in a fragmented market, and even a barrier to entry for slower-moving rivals.

✅ Actions for Founders & Operators

  • Map your AI surface: Determine if and how your product touches GPAI or high-risk AI.

  • Audit ready-or-not: Can you explain your training data? Do you have policies for incident logging?

  • Explore sandboxes: Consider EU sandboxes for compliance workflows and market access.

  • Plan for future waves: High-risk mandates come next year—now is the time to lay groundwork.

💬 Final Insight

The EU AI Act isn’t just a legal milestone—it’s a forcing function for product discipline. Founders who treat compliance as a checkbox will scramble. Those who treat it as a strategic foundation will build stronger, more future-proof companies.

Ignoring deadlines and geography is no longer an option. If you're building AI today, Europe is already at your doorstep.

Find out why 1M+ professionals read Superhuman AI daily.

In 2 years you will be working for AI

Or an AI will be working for you

Here's how you can future-proof yourself:

  1. Join the Superhuman AI newsletter – read by 1M+ people at top companies

  2. Master AI tools, tutorials, and news in just 3 minutes a day

  3. Become 10X more productive using AI

Join 1,000,000+ pros at companies like Google, Meta, and Amazon that are using AI to get ahead.