buildConfirmed6 publishers Google shipped two text-to-speech models with the same per-line performance control and put voice creation on only one of them, so the tier decision depends on whether a project has to invent a voice at all.
Perspective Coverage
7 publishers
- Builder
- Builder 64%
- Operator
- Operator 23%
- Investor
- Investor 13%
Reality
- Evidence62
- Adoption30
- Hype gap+25
- Incentives58
- Confidence66
buildConfirmed3 publishers ElevenLabs released Eleven v4 Turbo for voice agents, reporting medians of about 100 ms inference latency and 150 ms to first speech. Moving an existing agent takes more than a model ID swap, since older voice clones need retraining and SSML break tags no longer work.
Perspective Coverage
3 publishers
- Builder
- Builder 45%
- Operator
- Operator 33%
- Investor
- Investor 22%
Reality
- Evidence50
- Adoption
- Not scored
- Hype gap+25
- Incentives70
- Confidence60
Reporters asking Synthesia the basics can now talk to an avatar of its corporate affairs head
Synthesia, valued at $4 billion, has an interactive avatar of its corporate affairs head answering common press questions about the company. Comms leads thinking of doing the same need a rule for what a synthetic stand-in may say to outsiders.
Reality
- Evidence35
- Adoption25
- Hype gap+10
- Incentives65
- Confidence40
Universal Music and ElevenLabs announced a multi-year deal on September 10 for a remix and vocal platform still in development, with no launch date, no price and no participating artist named so far.
Perspective Coverage
3 publishers
- Builder
- Builder 23%
- Operator
- Operator 40%
- Investor
- Investor 37%
Reality
- Evidence58
- Adoption15
- Hype gap+30
- Incentives62
- Confidence60
Gemini 3.8 Flash TTS and Flash-Lite TTS arrive with matching APIs, more than 2,000 stock voices and a path from a 30-second sample to a replica voice. Securing the speaker's permission is the customer's job.
Reality
- Evidence34
- Adoption20
- Hype gap+32
- Incentives72
- Confidence55
buildOne report1 publisher VoxCPM2 trades the per-character speech bill for a GPU and a base-URL change
OpenBMB's 2B-parameter model is Apache-2.0, speaks 30 languages, and serves through vLLM's OpenAI-compatible /v1/audio/speech. The real-time factor quoted for it was measured on a different backend than the one the project recommends for production.
Reality
- Evidence38
- Adoption
- Not scored
- Hype gap+38
- Incentives42
- Confidence40
The FBI counted $893 million in AI-linked fraud losses in its first year of tracking them
The bureau's first year of counting AI-flagged complaints puts the losses on consumer money flows, where cheap voice cloning meets instant payment rails and any fix has to live inside the product itself.
Reality
- Evidence48
- Adoption62
- Hype gap+12
- Incentives58
- Confidence55
FBI logged 22,000 AI-linked fraud complaints in its first year of counting them
The bureau's 2025 total of roughly $893 million comes only from victims who filed and from cases where investigators could identify the AI, and two of its subtotals overlap enough to exceed it. It is a floor.
Reality
- Evidence55
- Adoption58
- Hype gap+15
- Incentives60
- Confidence45
Universal's multi-year deal with ElevenLabs covers product development as well as rights, though the announcement does not name the participating artists, set a launch date or disclose terms, so the roster gets decided after signing.
Perspective Coverage
5 publishers
- Builder
- Builder 23%
- Operator
- Operator 36%
- Investor
- Investor 41%
Reality
- Evidence57
- Adoption17
- Hype gap+34
- Incentives74
- Confidence66