Megumi Ogata, the veteran Japanese voice actor behind roles including Neon Genesis Evangelion, said AI voice cloning is “heartbreaking” and urged Japan to intervene with clear rules. Speaking to TBS News Dig, Ogata also said she has a hard time accepting why anyone would generate content that makes her “say things that I don’t want to say as they please” and then sells it as if it were her voice.

The target is not an abstract fear. Ogata pointed directly at the AI impersonator ecosystem on video platforms like TikTok, calling it an issue she “cannot turn a blind eye to.” She framed her voice as part of her identity, so the harm is not just technical. It is personal, reputational, and potentially commercial, because cloned audio can be packaged and distributed as if the actor authorized it.

This is happening while the Japanese government is “considering a legal framework similar to likeness rights” to address infringement. In other words, Japan is trying to shift from loose norms and platform moderation toward something closer to enforceable content ownership and consent expectations. For executives, this matters because it signals where regulators may land: not on “AI is bad,” but on “AI usage must respect identity rights.” The voice, in this framing, behaves like likeness, not like generic sound.

Ogata’s comments put weight behind a position other voice performers have raised. The source notes a similar stance from Wes Johnson, a frequent Bethesda collaborator, who said in a tweet that “anyone trying to create a mod using an actor's voice via AI without consent knows they are wrong.” Veteran videogame actor Jennifer Hale has also been vocal about studios taking responsibility for AI voice use, and at the 2024 BAFTAs, multiple members of Baldur's Gate 3's cast criticized AI voice cloning.

What is easy to miss is the weird boundary line these speakers often draw. The source highlights a “professional/amateur divide.” It is not just about big studios building tools. Ogata, Johnson, and Baldur's Gate 3 narrator Amelia Tyler also singled out modders and content creators as part of the problem. That matters because many companies and investors assume the biggest risk sits with formal production pipelines, where contracts and licensing can be audited. But the market reality is that audio can be produced and distributed by smaller actors, then amplified by platforms. That creates a compliance problem that is less about your internal workflow and more about how your content ecosystem, data sources, and distribution channels interact with unauthorized use.

For boards and product leaders, the second-order implication is liability and brand risk, not just PR. If regulators build a likeness-rights model, consent and authorization become the central currency. That can reshape how companies procure training data, how they negotiate voice talent agreements, and how they respond when third parties impersonate performers using AI. Even when a company is not the one generating the clone, an enforcement regime that recognizes identity rights can still drive demand for takedowns, proof of consent, and platform cooperation. In a worst case, a studio could find itself pulled into disputes it did not originate, simply because it benefits from an ecosystem that allows imitation.

There is also a creative-rights tension here that will feel familiar to anyone tracking IP law. Voice talent sits at the intersection of art and personal identity. The source underscores the creep factor even for “seemingly innocent” content like shitposts. That is not just moral language. It is a signal of how consumers and performers interpret the intent behind cloning, which can influence how quickly lawmakers and courts treat these works as harmful misuse rather than “transformative” experimentation.

Finally, this story is not just about Japan. Voice cloning is already used globally, and the regulatory question is whether enforcement will be localized or harmonized. Japan calling itself a “content superpower,” as Ogata did, hints at a strategy: set rules, export them as policy, and reduce friction for legitimate creators. For executives across media, games, and platforms, the strategic stake is simple. If you move fast without consent-grade governance, you can end up in a world where voice becomes a regulated identity right, not a flexible asset. And once rules harden, last-minute cleanups are expensive, slow, and reputationally loud.