Will Fastly disclose a quantified AI/agentic traffic contribution as a percentage of total traffic or revenue by Q2 2026 earnings?

Management declined quantification at Q4 2025 -- the single most favorable disclosure moment given the 64-88% stock surge driven partly by AI narrative. They had maximum incentive to validate the AI bull case and chose not to. The most parsimonious explanations: (a) AI contribution is immaterial and quantifying would deflate the narrative, or (b) CDN traffic attribution to AI agents versus traditional traffic is technically infeasible. Either explanation suggests non-disclosure persists through Q1 and Q2 2026 calls. The 19 insider sell transactions with zero buys suggests insiders are not confident enough to buy even as AI narrative inflated the stock. The broad resolution criteria (including customer count) provides a narrow path but management's revealed preference is clearly for qualitative framing.

Industry dynamics strongly favor continued non-disclosure. All CDN providers benefit from AI traffic growth -- Moat Mapper found no evidence of Fastly-specific AI capture. If Fastly quantified AI traffic and it proved immaterial (say <5% of traffic), the 20-30% of stock move attributed to AI narrative overlay would likely reverse. The Edge Compute segment where AI workloads are most measurable is only 4% of revenue ($21.1M FY 2025) -- quantifying this would actually undermine the bull case. Management's three qualitative AI categories (agentic traffic, compute workloads, AI products) are designed to maintain narrative optionality. The asymmetry of disclosure risk is stark: quantify and reveal immateriality = stock decline vs. keep vague = maintain narrative premium.

Two earnings calls remain (Q1 2026 in May, Q2 2026 in late July), providing two disclosure windows. The resolution criteria are notably broad: traffic percentage, revenue dollar amount or percentage, AI customer count, or AI-specific ARR all qualify. An AI customer count or AI-specific product ARR (for bot mitigation, RSL protocol) is substantially easier to disclose than parsing CDN traffic attribution. If Fastly launches a new AI product pricing tier, they might disclose customer traction as a go-to-market metric. This breadth of qualifying disclosures marginally increases probability. However, the committee finding that 4/5 lenses independently flagged the AI quantification gap suggests this is a structural reluctance, not an oversight. Competitive pressure from peers disclosing AI metrics could force Fastly's hand but no major CDN competitor has done so yet.

Management had the perfect opportunity to quantify AI traffic on Q4 2025 earnings and chose not to, despite a 64-88% stock surge partly driven by AI narrative. If they didn't quantify at maximum incentive, they won't at Q1 or Q2 2026. The AI narrative serves them better unquantified -- it allows market to project whatever AI exposure they wish onto the stock. Edge Compute is only 4% of revenue and AI-specific products (bot mitigation, RSL) are nascent. Insiders are unanimously selling (19 sells, 0 buys). The broad resolution criteria provide a narrow path via customer count disclosure but management's revealed preference strongly favors qualitative framing over quantitative commitment.

Against the bearish consensus: the resolution criteria are quite broad. 'AI customer count' is a much lower bar than 'AI traffic as percentage of total.' Companies frequently disclose customer counts for new product lines as a go-to-market credibility signal. Fastly's AI bot mitigation product and RSL protocol could generate disclosable customer metrics by Q2 2026 -- especially if they want to demonstrate AI product-market fit. Two full earnings cycles (Q1 May, Q2 July) is significant runway. If Cloudflare or Akamai disclose AI traffic metrics first, competitive pressure would force Fastly to respond. However, the committee's finding that management 'may have strategic incentive to NOT quantify' is compelling, and the disclosure risk asymmetry favors continued vagueness. This is a contrarian position acknowledging the broad resolution criteria provide more paths to YES than surface analysis suggests.

Four of five lenses independently flagged AI as unquantified E1 evidence -- this is the single most under-evidenced claim in the bull thesis. Insiders are unanimously selling with 661K shares disposed in 3 months. Management declined at the optimal moment. CDN traffic attribution is technically difficult -- AI agent HTTP requests look like regular requests. The stock move decomposition estimated 20-30% of the surge was AI narrative overlay, meaning management has financial incentive to preserve the narrative via ambiguity. Even the broad resolution criteria don't change the fundamental calculus: management will quantify only if forced to or if the numbers are impressively large. Neither condition is likely given the evidence.

Management avoided quantification at highest-incentive moment. All 19 insider transactions are sells, zero buys. 4/5 lenses flagged the gap. CDN traffic attribution is technically difficult. Strong NO lean. The broad resolution criteria (customer count, ARR) provide a narrow path but management behavior at Q4 2025 is the strongest signal pointing to continued non-disclosure.

Broad resolution criteria provide multiple paths to YES: traffic percentage, revenue percentage, customer count, or ARR figure. Two earnings calls remain before resolution. AI product launches (bot mitigation, RSL protocol) could generate customer count disclosures. But management's Q4 2025 behavior -- declining to quantify at the most favorable moment -- is the dominant signal. Probability modestly above base but still well below 50%.

No quantification at best moment, insider selling pattern (19 sells, 0 buys, 661K shares), technical attribution challenges for CDN traffic. The Myth Meter classified this as DIVERGING with E1 evidence only. Management's incentive structure strongly favors keeping AI narrative qualitative rather than quantitative. Resolution criteria breadth provides marginal upside path but not enough to materially shift the assessment.