People Inc. pivots to blocking all AI scrapers by default

People Inc.’s shift to a default-deny scraping posture reflects a hardening stance among premium publishers who refuse to provide free training data for LLMs. By funneling access through controlled marketplaces like Microsoft’s or direct deals with OpenAI, the company is attempting to standardize a 'pay-per-inference' or flat-fee revenue model. This sets a precedent for the broader streaming and digital media ecosystem: content is no longer a free resource for AI search, but a licensed asset with measurable unit value. Watch for whether more mid-tier publishers adopt similar 'default-deny' technical defenses through infrastructure providers like Cloudflare to force licensing negotiations.

The transition to 'default-deny' scraping strategies comes as major publishers increasingly seek to replace the voluntary robots.txt protocol with enforceable technical barriers. Per Search Engine Journal (June 2026), both Reuters and Time recently adopted similar allowlist-only policies, mirroring the move by People Inc. and The Atlantic. This shift is driven by data from platforms like Tollbit showing that roughly 30% of AI bot scrapes ignore robots.txt instructions. By moving to a default block, People Inc. reportedly identified more than 30,000 unauthorized user agents attempting to crawl its properties, a significant jump from the 2,100 it previously tracked on its manual blocklists.

Financially, these licensing agreements have become a critical revenue stream for legacy media groups. Per internal reporting and The Wall Street Journal (October 2025), News Corp’s landmark partnership with OpenAI is worth an estimated $250 million over five years, establishing a valuation benchmark for the industry. Similarly, People Inc.’s parent company, IAC, disclosed in November 2025 that the publisher had secured both a lump-sum arrangement with OpenAI and a usage-based deal with Microsoft. This dual-model approach allows publishers to capture value both from model training and real-time retrieval in tools like ChatGPT and Microsoft Copilot.

Simultaneously, publishers are deploying more aggressive cybersecurity tactics to protect proprietary data. According to Cloudflare (October 2025), new tools like 'AI Labyrinths' are being used to trap non-compliant scrapers in infinite loops of decoy data, while the IETF’s AI Preference Working Group works to formalize technical standards for AI content monetization. While 79% of top news sites now block at least one AI training bot, the emergence of 'pay-per-use' marketplaces suggests the industry is moving toward a highly fragmented web where premium content resides behind a layer of AI-compatible licensing filters.