Clear2Connect Coalition Strongly Opposes FCC Decision on Captioning with Automatic Speech Recognition

Article Featured Image

The Federal Communications Commission's (FCC) decision to conditionally approve a company to provide Internet Protocol Captioned Telephone Service (IP CTS) using only automatic speech recognition (ASR) technology raises concerns about whether the FCC is sanctioning services that could put veterans, the elderly, and other citizens at risk, the Clear2Connect has charged.

The FCC determined in June 2018 that ASR is permissible to deliver IP CTS based on a study conducted by the MITRE Corporation. The Clear2Connect Coalition has questioned the validity of that study because it was conducted in optimal lab conditions not representative of real-life telephonic communication, was not peer-reviewed, had a small sample size, and used varied testing methods for different technologies, rendering comparisons invalid. The latest;MITRE testing report was dated May 1.

ASR-only service cannot be determined to meet the requirement of functional equivalence guaranteed by the Americans with Disabilities Act (ADA) based on this flawed study, the organization claimed, noting that if the FCC continues to prematurely approve use of ASR technology, people who are deaf or hard of hearing will lose their ability to access accurate and reliable communication services.

The Clear2Connect Coalition is a group of 27 disability and veteran advocacy organizations from across the United States working together to protect the right for Americans with hearing loss to communicate via IP CTS. The coalition has consistently expressed concerns with approving ASR-only providers of IP CTS based on the MITRE or similarly flawed studies, noting that ASR-only services must be evaluated based upon real-life scenarios before approving them for consumer use.

Vietnam veteran and Clear2Connect Coalition member Fred Williams' experiences with IP CTS underscore the vital importance of human call assistants. "On all my phone calls, you can tell when a human is correcting something," he said in a statement. "Text will disappear, then reappear. It starts garbled, but a couple seconds later, a new one will come in improved. I can tell you that, without those corrections, huge portions of all my phone calls wouldn't make sense."

SpeechTek Covers
for qualified subscribers
Subscribe Now Current Issue Past Issues