Forrester says AI now handles 19% of inbound contact-center calls, up from 6% in 2024, and the errors are harder to spot than in text chat.
AI voice bots now handle about one in five inbound customer service calls. Forrester says the share reached 19% in 2026, up from 6% in 2024. The catch, according to a CustomerThink piece, is that companies are adding bots faster than they can check them. When a bot gets something wrong on a live call, the mistake can vanish.
Voice is natural, fast and familiar, and many customers prefer it to text for quick tasks. The money is following. Market.us projects the voice AI agent market will reach $47.5 billion by 2034. Gartner forecast that conversational AI will cut contact center labor costs by $80 billion in 2026.
A text chatbot leaves a paper trail. Customers can re-read the answer, screenshot it or forward it. A phone call leaves much less for the customer to point to. That makes it harder for a company to see what went wrong and harder to prove the bot behaved correctly.
Most phone bots also work in three steps. They turn speech into text, have an AI write a reply, then turn the reply back into speech. Context, tone or intent can get lost along the way. Background noise or accents can cause mishearing. The bot may interrupt people, miss frustration, or give a wrong answer in a confident voice. Newer bots that work directly with speech have only just arrived from OpenAI, Google and xAI.
The article gives three examples of bots that kept going after a serious error:
Problems can also sit outside the bot. A handoff to a human may fail if the agent's computer is slow to load the customer's history, or if the connection drops. Context may not carry over, so customers have to repeat themselves.
For any business using or buying a phone bot, the advice is to test beyond the easy, happy-path calls. Try messy ones with noise, accents, changed requests and upset callers. Keep testing after launch, not just before it. Set clear rules for when the bot should ask a question or pass the call to a person. Make sure a customer can reach a human quickly, and that the human sees what was already said.
Voice bots are likely to keep growing. The article argues that trust will depend on ongoing monitoring across the whole customer journey, not a one-time check at launch. It mentions five governance practices but does not spell them out in the text available.
If a business uses a phone bot, test it on messy real calls and make sure customers can reach a human fast.
Otto, a new customer-tracking app for field sales teams, lets reps speak their notes instead of typing them. Price and launch details are not yet known.
Read more: Otto launches a sales app you talk to instead of typingGoogle Home can now check your home Wi-Fi to decide if you are there, which may fix lights and thermostats that misfire when you come and go.
Read more: A Wi-Fi setting makes Google Home know when you're really homeA writer gave Claude control of his rooted LG TV and fixed slow startup, uneven app volume and a dull screensaver. The privacy gains are less certain.
Read more: A writer used Claude to fix his LG TV's slow startup and loud apps
