Reddit says it has to cut back access to ‘Old Reddit’ because of AI bots

Reddit will further restrict access to its Old Reddit interface in the coming months, requiring users to be logged in and to have used the legacy site within the prior six months. The measure is intended to curb scraping and automated traffic, while Reddit says the eligibility window covers the vast majority of existing Old Reddit users. The change could modestly increase friction for legacy users but supports the company's control over platform access and bot activity.

Analysis

The direct financial effect is immaterial, but the policy improves the quality of RDDT's logged-in audience and reduces low-value infrastructure demand. If enforcement meaningfully lowers automated collection, Reddit can preserve scarcity around its conversational data and strengthen leverage in licensing negotiations with AI developers; the relevant upside is higher data-licensing durability and lower serving costs, not near-term advertising revenue.

Near term, markets may read any traffic decline as engagement deterioration, particularly if third-party measurement shows lower pageviews before logged-in daily users stabilize. That creates a 1-3 month watch point: a decline in anonymous traffic is constructive only if DAUq, session depth, ad impressions per user, and API/data revenue remain intact. A material fall in search-referral visibility or community contribution rates would instead signal that friction is impairing the content flywheel.

The non-obvious risk is substitution: restricting the legacy interface can push technically sophisticated, high-intent users toward decentralized forums, Discord, X, or niche communities. Those users disproportionately create searchable long-tail content that supports Reddit's Google referral funnel and AI-data corpus. Over 6-18 months, the key question is whether reduced scraping protects monetizable proprietary data faster than access friction erodes content supply.

Consensus may overfocus on bot prevention as universally positive. Scraper traffic can inflate costs, but it also helps distribute Reddit content across the web and can indirectly reinforce relevance; a hard access gate could lower the observed data footprint sought by AI partners. This is not independently verifiable as a revenue catalyst until management quantifies API utilization, licensing contract value, or infrastructure savings.

Key Decisions for Investors

• No standalone RDDT trade on this announcement; impact is below the threshold for a position absent evidence of a measurable change in logged-in engagement or data-licensing economics.
• Maintain a 1-3 month RDDT watch: become constructive only if third-party traffic softens while reported DAUq and ad impressions per DAU remain stable or improve, indicating successful conversion from anonymous to authenticated usage.
• Treat a sustained deterioration in search referrals, posting/comment volume, or DAUq at the next earnings print as thesis falsification; this would imply the access friction is damaging the content and discovery flywheel rather than removing non-monetizable automation.
• For existing RDDT longs, frame any rally tied solely to 'anti-scraping' as an opportunity to trim unless accompanied by disclosed infrastructure savings or incremental data-licensing revenue; monetization proof, not policy language, should drive multiple expansion.