# What Did ElevenLabs' CEO Actually Say About Bot Disclosure?

TechCrunch's Connie Loizos interviewed ElevenLabs CEO Mati Staniszewski in a piece published September 24, 2026. According to TechCrunch, Staniszewski said businesses should probably tell customers they're talking to a bot — at least until getting a machine is what everyone expects anyway. That qualifier matters more than the disclosure endorsement itself. Staniszewski is not arguing for permanent transparency. He is arguing for transitional transparency — a position that gives ElevenLabs cover with regulators now while preserving the option to deprioritize disclosure as a feature once consumer expectations shift. TechCrunch reported the company is now valued at $22 billion, which reframes the CEO's comments as investor-facing risk management rather than a product philosophy. The disclosure question is not abstract for ElevenLabs. Its technology sits behind customer service calls, which means every deployment carries an implicit consent question that the company does not directly control. The CEO can recommend disclosure; he cannot enforce it across enterprise customers optimizing for call resolution rates.

# Why Does a $22 Billion Valuation Change the Disclosure Calculus?

According to TechCrunch, ElevenLabs is now reportedly valued at $22 billion. That figure — reported, not confirmed by the company — creates a specific set of pressures that a smaller private company would not face. At that valuation, ElevenLabs needs either a public offering or a strategic acquisition to deliver returns to investors. Both paths require auditable revenue quality, which means enterprise contracts that survive regulatory scrutiny. Bot disclosure is not just an ethics question at this scale; it is a diligence question. Any IPO prospectus that lists customer service deployments as a core revenue driver will invite questions about whether end users know they are interacting with AI. TechCrunch reporter Connie Loizos has covered Silicon Valley since the late 1990s, per her TechCrunch author bio, and her interview frames Staniszewski's comments as part of a broader conversation about margins and IPO timing. That framing is telling: the CEO is managing multiple stakeholder narratives simultaneously — regulators, enterprise buyers, and future public market investors.

# How Do ElevenLabs' Margins Compare to the Rest of Voice AI?

ElevenLabs has not publicly disclosed gross margins. The TechCrunch interview references margins as a topic of discussion but does not provide a specific figure. What is verifiable is the valuation: $22 billion, reported by TechCrunch. Voice synthesis at scale carries infrastructure costs that text-based AI does not. Real-time, low-latency voice generation requires GPU capacity that scales with concurrent call volume, not just training runs. That structural difference means ElevenLabs' margin profile is more sensitive to usage growth than a pure software API business. The comparison table below reflects what is publicly known versus what is inferred. I have labeled inferences explicitly.

Company	Reported Valuation	Disclosure Stance	Margin Profile (Inferred)	IPO Signals
ElevenLabs	$22B (TechCrunch, Sept 2026)	Transitional disclosure — recommended, not mandated	Usage-sensitive; voice inference costs scale with concurrent calls	CEO discussing timing publicly
OpenAI	Not comparable (multi-product)	No specific voice disclosure policy announced	Diversified across text, voice, video	No IPO signal
Google (Cloud Voice)	Parent company public	Subject to existing Google AI disclosure guidelines	Bundled into Cloud margins	Already public
Enterprise contact-center vendors	Varies widely	Customer-configurable disclosure	Software margins with voice API pass-through	Not applicable
Verdict	ElevenLabs leads on valuation transparency	ElevenLabs is hedging; competitors are silent	Margin risk is highest for pure-play voice	ElevenLabs is the only pure-play voice company signaling IPO readiness

# What Do AI Labeling Rules Require Right Now?

The EU AI Act's transparency obligations for AI systems interacting with humans took effect in August 2026, requiring disclosure unless it is obvious to a reasonably informed person. That standard — "obvious to a reasonably informed person" — is exactly the ambiguity Staniszewski is navigating. TechCrunch reported his position as conditional: disclose now, reassess later. The US has no equivalent federal labeling mandate as of September 2026, though the FTC has brought enforcement actions against deceptive AI claims under existing authority. State-level bills in California and Colorado have advanced disclosure requirements for AI-generated communications, but none specifically target real-time voice agents in customer service. This regulatory patchwork explains why ElevenLabs is not issuing a hard mandate to customers. A hard mandate would satisfy EU regulators but potentially alienate US enterprise buyers who see disclosure as a conversion-rate killer. A soft recommendation preserves flexibility in both jurisdictions.

# Who Wins and Who Loses If Disclosure Becomes Standard?

Enterprise contact-center vendors that already offer configurable disclosure win. They can sell compliance as a feature. Pure-play voice API companies like ElevenLabs lose relative flexibility because they do not control the end-user experience — their customers do. Consumers win in the short term if disclosure is enforced, because they can choose to escalate to a human. In the long term, if Staniszewski's prediction holds and machine interaction becomes expected, disclosure becomes a formality rather than a meaningful signal. The biggest loser in a mandatory-disclosure regime would be any voice AI company that has optimized its pricing around high-volume, low-friction deployments. Those deployments assume users do not hang up when they learn they are talking to a bot. If disclosure increases hang-up rates even 5%, the unit economics of those contracts change materially.

# Predictions

1. ElevenLabs will file for an IPO or announce a strategic acquisition process before Q4 2027, driven by the $22 billion valuation pressure reported by TechCrunch and the need to provide liquidity to early investors. 2. The EU AI Office will publish implementing guidance on AI voice disclosure before June 2027, and at least one US state — likely California — will pass a voice-agent-specific disclosure law before the end of 2027. 3. At least one major ElevenLabs enterprise customer will publicly disclose that it uses AI voice agents without verbal disclosure, triggering a consumer protection inquiry or FTC warning letter before Q3 2027.

# Article Summary

• ElevenLabs CEO Mati Staniszewski's conditional disclosure stance is a regulatory hedge, not a transparency commitment — and the $22 billion valuation makes that hedge a market signal.
• The company's margin profile is structurally more sensitive to usage growth than text-based AI competitors because real-time voice inference scales with concurrent call volume.
• EU AI Act transparency obligations, effective August 2026, create the regulatory backdrop for Staniszewski's comments, but the "reasonably informed person" standard remains undefined.
• Enterprise contact-center vendors are positioned to win if disclosure becomes mandatory because they can sell compliance as a feature; pure-play voice API companies lose flexibility.
• The IPO window ElevenLabs is eyeing depends on converting disclosure from a cost center into a compliance product before regulators force the issue.