Meta’s new commercial terms will allow the scraping of news sites: experts

Original Article ↗ Hacker News Discussion ↗

Meta’s new commercial terms for business accounts, which grant it broad rights to scrape and reuse content from linked websites (including news sites) for AI training, are raising alarms about consent, copyright, and power imbalances. Commenters question whether such clickthrough agreements are enforceable, how they interact with third‑party licenses (like stock images or syndicated content), and whether small publishers could be left legally exposed while Meta shields itself. Many see the move as an attempt to normalize large‑scale data harvesting and bypass paywalls or robots.txt, deepening calls for antitrust action and stricter regulation, especially outside the U.S.

Overall reaction to Meta’s new terms

  • Many see the change as aggressive “legal engineering” giving Meta broad rights to scrape and reuse business content, especially for AI.
  • Tone is largely negative toward Meta as a company (data exploitation, societal harm), though a minority notes some useful products (Marketplace, Groups) while others counter that these are inferior, lock‑in replacements.

Legality of scraping and clickwrap / ToS

  • Several question whether such a broad, one‑sided term in a clickwrap or business contract would be enforceable, calling it “unconscionable.”
  • Others argue courts already treat “click accept” like signing paper contracts; users and businesses are assumed to have read terms.
  • Debate on ToS changes: some want automatic opt‑out until explicitly accepted; others say most people do not care and the system can’t “babysit.”
  • Strong disagreement over whether scraping public sites is inherently legal; some liken it to public observation, others point to copyright (e.g., Disney assets, music in bars) as clear limits.

Third‑party rights and liability

  • Multiple commenters note publishers often don’t own full rights (e.g., stock images, licensed content) and cannot legally sublicense to Meta.
  • View 1: putting content publicly online implies you can authorize such use; if not, you likely shouldn’t host it.
  • View 2 (stronger in thread): sublicensing rights are frequently restricted; Meta’s consent clause can’t conjure rights “from thin air.”
  • Explanation offered: Meta may rely on indemnification—publishers attest they have rights; if Meta is sued, it pushes liability and costs onto them.
  • Others think this mainly serves as legal cover for behavior Meta is already engaging in, especially AI training.

Business, antitrust, and practical impact

  • Some expect businesses to leave Meta or see an antitrust angle (leveraging ad dominance to obtain training data); others note businesses rarely quit despite past abuses.
  • Small operators report already blocking Meta bots, citing deceptive user‑agents and heavy crawling; say they will ignore the new ToS.
  • Comparisons to Google/Bing: some see little difference; others highlight ignoring robots.txt and conflating link‑preview crawlers with AI scrapers.

Resistance and counter‑measures

  • Suggestions include:
    • Serving alternate or misleading content to Meta’s crawlers.
    • Imposing site‑level ToS that charge or forbid Meta scraping, though enforceability is seen as legally unclear.
    • Using legal tools like copyright registration to pursue higher damages if infringement occurs.

Jurisdiction and regulation

  • Unclear how these terms fare under EU, Australian, or Canadian law; some cite “unfair contract terms” regimes and recent media bargaining laws as potential friction but no consensus emerges.