๐The Next Web (TNW)โขStalecollected in 34m
UK government partners with ElevenLabs for accessible public services
๐กSee how governments are adopting generative voice AI to scale public services and improve accessibility.
โก 30-Second TL;DR
What Changed
Memorandum of Understanding signed between UK government and ElevenLabs
Why It Matters
This deal sets a precedent for government-private sector collaboration in deploying generative AI for public infrastructure. It validates ElevenLabs' technology for large-scale, sensitive government applications.
What To Do Next
Explore ElevenLabs' API documentation to understand how to implement low-latency, multilingual voice synthesis in your own accessibility-focused applications.
Who should care:Developers & AI Engineers
Key Points
- โขMemorandum of Understanding signed between UK government and ElevenLabs
- โขFocus on making public services accessible in any language and voice
- โขInitiatives include safety research and building local AI talent
๐ง Deep Insight
Web-grounded analysis with 23 cited sources.
๐ Enhanced Key Takeaways
- โขThe Memorandum of Understanding (MoU) between the UK government and ElevenLabs is a three-year partnership specifically designed to research the societal and security implications of AI voice technology, including how effectively people can identify AI-generated voices and how this knowledge influences their behavior.
- โขThis partnership expands upon an existing research collaboration between ElevenLabs and the UK AI Security Institute (AISI), which was initially announced in February 2026.
- โขThe UK government's agreement with ElevenLabs represents its fifth such partnership with a frontier AI company, following previous MoUs signed with other major AI entities like OpenAI and Google DeepMind.
- โขThe initiative aims to enhance accessibility for diverse groups, including individuals with visual impairments, low literacy, low digital confidence, and those from linguistically diverse communities, with a specific mention of supporting Welsh-language services.
- โขElevenLabs had launched its 'ElevenLabs for Government' initiative just a week prior to the MoU signing, offering 24/7 multilingual services to public sector organizations, with the Government of Ukraine already an early adopter.
๐ Competitor Analysisโธ Show
| Company | Key Features | Latency / Benchmarks | Pricing Model (General) |
|---|---|---|---|
| ElevenLabs | Human-like voice models, instant/professional voice cloning, voice design, 70+ languages (Eleven v3), 3000+ community voices, STT (Scribe v2, 90+ languages), multi-speaker dialogue. | ~75ms for Flash models (real-time), 400-800ms TTFB for async content. | Freemium, paid plans from $5/month, enterprise pricing. |
| Deepgram Aura-2 | Built for real-time enterprise conversations, unified STT+TTS architecture. | 90ms optimized TTFB, 200-250ms full conversation latency. | Enterprise-focused. |
| OpenAI TTS | Developer-friendly integration. | Not specified. | Competitive pricing. |
| Google Cloud Text-to-Speech | Advanced AI voice synthesis, vast language/accent range, enterprise ecosystem. | Not specified. | Cloud-based, usage-based. |
| Amazon Polly | Lifelike voices, 60+ languages, flexible. | Not specified. | Pay-as-you-go. |
| Fish Audio | High-quality (ranked #1 in TTS-Arena blind tests), open-source option, 2M+ community voices, emotion tags. | Quality rivals ElevenLabs. | Pro plans from $9.99/month, API $15/million characters (80% cheaper than ElevenLabs). |
| Chatterbox (Resemble AI) | MIT-licensed open-source, outperformed ElevenLabs in blind tests (63.8% preference), 5-10s voice cloning, 23 languages, emotion control, watermarking. | Under 150ms latency for paid API. | Free (open-source for commercial use), paid API. |
๐ ๏ธ Technical Deep Dive
- Speech Synthesis Models: ElevenLabs offers several models tailored for different use cases:
- Eleven v3: The flagship model, providing the highest fidelity, rich emotional expression, and support for over 70 languages. It excels in character discussions and audiobook production, supporting natural multi-speaker dialogue and inline audio tags for emotional control.
- Eleven Multilingual v2: Designed for lifelike, consistent quality speech synthesis across 29 languages, best for voiceovers and content creation, and stable for long-form generations.
- Eleven Flash v2.5: An ultra-low-latency model (approximately 75ms) optimized for real-time applications and conversational agents, supporting 32 languages.
- Voice Generation Capabilities: The platform supports various methods for voice creation, including a voice library with over 3,000 community-shared voices, instant voice cloning, professional voice cloning for high-fidelity replicas, and voice design to generate custom voices from text descriptions.
- Emotional Context and Control: Models interpret emotional context directly from text input (e.g., "she said excitedly") and allow fine-tuning through voice settings like Stability and Similarity. Eleven v3 further enhances control with inline audio tags.
- Speech-to-Text (STT): ElevenLabs also provides a state-of-the-art STT model, Scribe v2, which offers highly accurate transcription across more than 90 languages. It includes advanced features such as speaker diarization (distinguishing multiple speakers), entity detection, and precise word-level timestamps.
- API Access: All models are accessible via API, with specific pricing and options for different use cases and latency requirements.
๐ฎ Future ImplicationsAI analysis grounded in cited sources
The partnership will accelerate the adoption of voice AI in UK public services.
The MoU explicitly focuses on integrating voice AI for accessibility and 24/7 multilingual services, building on existing government initiatives and ElevenLabs' 'for Government' offering.
ElevenLabs will likely enhance its AI safety features and research capabilities due to the UK government partnership.
A core pillar of the three-year MoU is dedicated to collaborative research into the societal and security implications of AI voice technology and developing safeguards.
The UK will become a leading hub for AI voice talent and research.
The partnership includes initiatives for building local AI talent and attracting international expertise in voice and audio AI.
โณ Timeline
2022
ElevenLabs founded by Piotr Dฤ
bkowski and Mati Staniszewski.
2023-01
ElevenLabs secured $2 million pre-seed funding and publicly released its beta platform.
2024-01-22
ElevenLabs raised $80 million in Series B funding, valuing the company at $1.1 billion.
2026-02-04
ElevenLabs raised $500 million in Series D funding, reaching an $11 billion valuation.
2026-02
ElevenLabs launched 'ElevenLabs for Government' and announced a research partnership with the UK AI Security Institute (AISI).
2026-06-08
UK government signed an MoU with ElevenLabs for accessible public services.
๐ Sources (23)
Factual claims are grounded in the sources below. Forward-looking analysis is AI-generated interpretation.
๐ฐ
Weekly AI Recap
Read this week's curated digest of top AI events โ
๐Related Updates
AI-curated news aggregator. All content rights belong to original publishers.
Original source: The Next Web (TNW) โ