Recent breakthroughs in AI are helping survivors communicate using their own unique sounds and intentions, even when speech is difficult.
1. Apple’s “Personal Voice” (iOS 17 & 26)
This is a game-changer for those whose speech may fluctuate or who are at risk of losing their voice.
- What it is: Your iPhone or iPad can create a “digital twin” of your voice.
- How it helps: Instead of using a robotic computer voice, you can type a message and your device will speak it out loud using your recorded voice.
- The Update: In the latest iOS 26 update, Apple has made this even easier. You now only need to record 10 brief phrases (which takes about a minute) to create a lifelike version of your voice.
- Where to find it: Go to Settings > Accessibility > Personal Voice.
2. Google’s “Project Relate” (Android)
Standard voice recognition often fails for people with aphasia or stroke-related speech differences. Google is training AI to “learn” non-standard speech patterns.
- Listen: It transcribes your unique speech into text in real-time (like live subtitles for your own voice).
- Repeat: It can repeat what you’ve said in a clear, synthesised voice—ideal for face-to-face interactions like ordering a coffee.
- Status: This is currently an Android-only “Beta” app. You can register your interest at sites.research.google/relate.
3. “Voiceitt” (The Independent Alternative)
If you don’t use Apple or Google products, Voiceitt is a standalone app that “translates” dysarthric or non-standard speech into clear, fluent speech.
- Personal Training: You “train” the app by repeating certain phrases. The more you use it, the better it understands your specific sounds.
- Integrations: It can be linked to smart home devices (like Alexa) so you can control your lights or music using your own voice.
4. AI Writing Assistants (The “Sheffield Method”)
University research (including recent work at the University of Sheffield) has shown that tools like ChatGPT or Grammarly can be powerful for aphasia survivors.
- Expansion: A survivor can type three keywords (e.g., “Garden, tea, sunny”), and the AI can expand that into a full, polite email: “I am sitting in the garden enjoying the sun with a cup of tea.”
- Control: The survivor stays in control of the idea, while the AI handles the structure.
🔒 A Note on Privacy: Is My Voice Safe?
When using these AI tools, you are sharing very personal information. It is natural to wonder where that data goes. Here is how your privacy is protected:
- On-Device Processing: The most secure tools, such as Apple’s Personal Voice, do not send your voice to a computer in the cloud. The “learning” happens entirely inside your own phone, so your voice stays in your pocket.
- Encryption: If you share your “Personal Voice” across your own devices (like your iPhone and iPad), the data is End-to-End Encrypted. This means it is scrambled into a code that only your devices can open.
- Biometric Locks: Most phones require Face ID, Touch ID, or a Passcode to use these features, preventing strangers from using your digital voice.
- You are in Control: With reputable apps like Voiceitt or Google Project Relate, you have the right to delete your data at any time and opt-out of “training” features that use your data to improve the AI.
✅ Jonathan’s Safety Checklist
Before you download a new AI app, look for these three things:
- “Privacy by Design” – Does the app mention privacy in its description?
- No Third-Party Sharing – Does it promise not to sell your data to advertisers?
- Local Storage – Does it keep your recordings on your phone rather than the internet?
Jonathan’s Advice: “AI isn’t here to speak for us; it’s here to give us the scaffolding to speak for ourselves. Think of your digital voice like your credit card details—as long as you keep your phone’s passcode private and stick to well-known apps, it is a very safe way to regain your independence.”