Grammarly’s CEO defends putting AI editorial suggestions into the voices of real writers (while noting it didn’t work very well)

Superhuman (formerly Grammarly) pulled its 'Expert Review' feature after eight months following backlash and a class-action lawsuit filed by journalist Julia Angwin; CEO Shishir Mehrotra apologized on a Decoder podcast. The feature framed AI-generated editing suggestions as inspired by named public figures (journalists, authors) with attribution links, prompting claims of impersonation and reputational/legal risk despite the company disputing the claims. Implication: limited near-term market impact but heightened litigation and regulatory scrutiny around AI attribution and name/likeness usage that could affect product strategy and trust.

Analysis

The immediate fallout from persona-style features creates a two-track market: firms that can offer permissioned, revenue-share persona licensing and provenance tooling will extract value; consumer-facing startups that treat names/voices as costless growth hacks will see higher compliance burn and concentrated legal tail risk. Expect incremental legal/compliance spend of 2–5% of ARR for startups that keep persona features live, and potential one-off legal/settlement hits in the $10–200M range for mid‑sized players if class actions proliferate over 6–24 months.

Incumbent cloud and enterprise vendors win optionality because they can bundle governance, identity, and provenance as upsell modules — that moves monetizable margin from ad/toll models to license/transaction fees (estimate: 5–15% take rate on verified‑persona monetization for platforms that broker creator relationships). Publishers and legacy creators gain leverage to demand revenue share or licensing fees, creating a renewed content premium: verified content channels could command 5–10% higher CPMs or subscription ARPUs within 12–36 months.

Catalysts to watch are regulatory guidance or court rulings on name/likeness and AI attribution (timing: 6–18 months), large settlements that set precedent, and adoption of standardized provenance/watermarking by major LLM providers (3–9 months). Contrarian angle: the market is overdiscounting the monetization path — permissioned persona ecosystems will create new, durable revenue streams for platforms that move quickly to sign creators, so this is as much a structural reallocation of rents as it is a litigation shock.

Key Decisions for Investors

• Long MSFT: Buy a 9–12 month call spread (size 2–5% portfolio) to capture enterprise re‑platforming into permissioned, auditable models; capped upside limits theta drag while preserving ~2–4x upside if commercial AI adoption accelerates post‑governance standardization. Hedge with 1% portfolio cash for event‑driven legal headline risk.
• Long GOOGL: Purchase 9–12 month calls or call calendar (size 1–3% portfolio) — Google’s control over search, ads, and model governance means it can sell provenance/attribution as a product; expect meaningful re‑rating if publishers sign revenue‑share deals within 12 months. Target 20–30% return vs ~10% downside in a slow adoption case.
• Long ADBE: Buy 12‑month calls (size 1–2% portfolio) — Adobe can bundle creator licensing, identity, and editing workflows; a successful creator‑licensing marketplace would expand margins and justify a 10–20% multiple expansion over 12–24 months. Keep position light; downside is limited if creative spending softens.
• Long NYT: Buy 6–12 month calls or accumulate shares (size 1–2% portfolio) — legacy publishers that can tokenize or license verified expertise stand to capture a disproportionate share of incremental AI output monetization; a successful roll‑out could add 5–10% to ARPU in 12–24 months. Risk: slow adoption and tech integration delays.
• Short ARKK (hedge): Buy 6–9 month put spread on ARKK (size 1–3% portfolio) to express skepticism on unprofitable, consumer‑facing AI startups being re‑rated while facing legal/compliance shocks; expect 20–40% downside potential in worst‑case hype unwind vs limited premia paid for puts. Close on first major regulatory clarification or large settlement precedent.