Audio & Speech

Today's Stories

1. 1

   AI voice-cloning fuels surge in social engineering attacks

   Scammers are increasingly using AI-powered voice-cloning and deepfake technology to imitate trusted contacts, send personalized phishing emails, and conduct coordinated attacks across email, phone calls, messaging platforms and social media simultaneously. These capabilities, once requiring specialized expertise, are now widely accessible through readily available tools. AI has dramatically accelerated the speed and scale at which social engineering attacks can operate. Attackers can generate hyper-realistic emails, clone voices, and create manipulated audio, images and videos tailored to specific individuals or teams in minutes—making it harder for employees to spot fraud before disclosing sensitive information or granting system access.

   The article recommends verifying any unexpected request involving credentials, approvals, payments or sensitive information through a separate trusted communication channel, following established business processes, and reporting suspicious communications promptly. Independently confirming requests can help prevent fraud and impersonation attempts.

2. 2

   Deepfake voices pose growing risk to diplomatic communications

   A Reddit discussion raises concerns about AI voice cloning targeting diplomats and government officials, noting that voice recognition alone no longer serves as reliable identity verification when realistic synthetic voices can be created. If adversaries can convincingly clone the voices of known contacts, officials may be tricked into sharing sensitive information or making decisions based on fraudulent calls, potentially undermining secure diplomatic and government communications.

   The discussion acknowledges uncertainty about how frequently deepfake calls are occurring in practice and whether officials have already begun changing verification protocols in response to the threat.

3. 3

   Wispr raises $280M for AI speech-to-text at $2B valuation

   Wispr AI, which makes Flow—a dictation tool that converts spoken words into clean text—raised $280 million in Series B funding at a $2 billion valuation, led by Menlo Ventures. The company also previewed Canto, its first proprietary AI model trained to isolate human speech in noisy environments. Canto reduces transcription error rates in loud settings from 30% to nearly 5%–10%, making dictation practical beyond quiet rooms—in cars, offices, or near traffic. This matters for mobile workers, but also for deaf and hard-of-hearing users who rely on AI transcription in real-world, multi-speaker settings. Wispr says people have written over 60 billion words on Flow and the tool is used by almost all Fortune 500 companies and over 10,000 enterprises.

   Canto is in preview. The company's total raised now stands at $361 million after this round, which included new investors Acrew, Forerunner, Goodwater, Peak XV, Together Fund, and PLUS Capital alongside existing backers.

4. 4

   Regal Voice AI agents now available via Five9's contact center platform

   Enterprise voice AI company Regal Voice has integrated its autonomous AI voice agents into Five9's contact center platform through Five9's AI Agent Connect program. The integration allows Five9 customers to trigger follow-up calls or texts from Regal's AI agents automatically after a call event, with call records and contact history flowing between the two systems. Regal's software can also score contacts for intent and route high-intent leads directly into Five9 outbound campaigns without manual intervention. The partnership expands Regal's reach into Five9's customer base of more than 3,500 enterprises across more than 140 countries. For contact centers in regulated industries like healthcare, insurance, and education—where Regal already operates—the integration removes manual queueing steps and lets AI and human agents start with full call history. Regal claims a 99.5% resolution rate without human handoff, so the pairing may reduce operational friction for enterprises managing high call volumes.

   Regal has raised $83 million to date, most recently $40 million in October 2024. Five9 reported strong results on April 30, with subscription revenue growth accelerating for a second consecutive quarter and the company raising its full-year outlook; the stock rose about 18% on those results. Both companies are signaling momentum in agentic contact center software.

5. 5

   VocalCode brings push-to-talk dictation to code editors, $4.99 one-time buy

   VocalCode, a desktop dictation tool, lets developers speak code and prose directly into editors, terminals, and chat windows using on-device speech recognition. The tool costs $4.99 as a one-time purchase (no subscription), includes a 30-day free trial, and works on up to 3 devices. Speech recognition runs entirely on your CPU—audio never leaves your computer—making it useful for developers who want privacy-preserving hands-free input. The tool supports Mandarin Chinese and 25 European languages via auto-detecting models, plus English, so teams across regions can use it without uploading speech data to the cloud.

   VocalCode requires macOS 11 or later on Apple Silicon, or Windows; the 25 European language options are model-supported previews pending per-language testing. After the initial model download, recognition works fully offline; revalidation of the paid licence happens online at least every 30 days.

6. 6

   AI gives nonspeaking opera singer expressive voice

   Jakob Jordan, a young man with autism and apraxia who cannot reliably speak aloud, starred in an opera called Sensorium Ex using a custom AI speech synthesis device designed by New York University professor Luke DuBois. The device samples Jakob's own disordered speech, infers 401 parameters that represent the acoustic architecture of his voice, and synthesizes new text through those parameters—allowing him to hear his thoughts spoken in a voice that sounds like him, with full expressive control via lidar sensors that let him speed up, slow down, and change pitch in real time. Traditional text-to-speech devices for people with communication disabilities (AAC devices) treat the synthesized voice as generic—monotone and emotionless. Jakob's system shifts the frame: instead of just outputting text, it gives him what users identified as a civil rights need—the right to inject emotional affect, to sound sarcastic or joyful or upset. Composer Paola Prestini built the opera around this insight, working with poet Brenda Shaughnessy (who has lived experience as a mother to a nonverbal child) to honor the disabilities written into the characters and create a framework where disabled performers could thrive.

   The device runs entirely on a local computer using lidar sensors (the same $6 technology that deploys car airbags) rather than cloud services, preserving privacy by eliminating third-party data exposure. Jakob continues to use a traditional text-to-speech device for everyday life; the AI voice is his choice for artistic expression and performance. DuBois notes that entering what you want to say remains slow and cumbersome—the team's next challenge.