Summary

• BT Business has launched AI from BT, beginning with an AI receptionist that handles calls, bookings, transcripts, and routine enquiries.
• BT-commissioned research found 65% of small-business owners see significant potential in AI, but 56% say they are not yet realising its full value.
• The product illustrates how established suppliers can package AI around defined operational problems instead of expecting smaller businesses to assemble their own systems.

Artificial intelligence is being packaged into increasingly ordinary business services, with BT Business launching an AI receptionist for UK small companies that want to automate telephone enquiries without replacing their existing phone systems.

AI Receptionist is the first product in a wider portfolio called AI from BT. It can answer calls continuously, respond to routine questions, schedule appointments, send confirmations, capture conversations, and produce transcripts, while businesses can keep their existing telephone number.

BT Business says the service works across UK fixed and mobile networks and can integrate with Microsoft Outlook and Google Calendar. Customers can provide their own business information to shape how the automated receptionist responds.

The launch targets a part of the market where enthusiasm for AI has not necessarily produced useful deployment. BT-commissioned research found 65% of small-business decision-makers believe AI could significantly benefit their organisation, while 56% say they are not yet realising its full value.

A defined job makes adoption easier to measure

Smaller businesses face a different implementation problem from large enterprises because they are less likely to employ dedicated data teams, AI architects, procurement specialists, or software engineers capable of connecting models safely to business systems.

Packaged applications can therefore provide a more practical route than asking a company to assemble automation from APIs, telephony infrastructure, models, workflow software, and security controls. An AI receptionist also gives the customer a recognisable task rather than an abstract promise of productivity.

BT’s research suggests that telephone handling remains sufficiently important to create a measurable case. Among the businesses surveyed, 67% said they miss at least one call on an average day that they believe could result in lost business, while 58% spend between 30 minutes and two hours each day answering customer calls.

The company has attached a headline estimate of £3.7 billion in annual revenue lost by UK small businesses through unanswered calls, although that figure is commissioned modelling rather than a direct measure of transactions that disappeared.

Censuswide surveyed 1,500 decision-makers at UK businesses with fewer than 250 employees between 23 and 28 September, and respondents estimated revenue they believed had been lost because calls went unanswered. BT assigned midpoint values to response bands, excluded “don’t know” answers, and annualised nine months of reported losses to produce the £3.7 billion figure.

That methodology makes the number useful as an indication of the perceived scale of the problem while leaving substantial uncertainty around the exact economic loss. Businesses cannot know with certainty whether every unanswered call would have converted into revenue.

Operational AI also carries operational responsibility

Automating telephone conversations creates more concrete governance requirements than giving employees access to a writing assistant because calls may contain names, contact information, appointment details, financial circumstances, health information, or other sensitive material depending on the business involved.

BT says customer data used by AI Receptionist remains private and is not used to train AI models, which addresses one concern for smaller companies that may have limited capacity to inspect the data practices of specialist AI vendors.

Businesses still need to decide where automation should stop, because scheduling an appointment or answering an opening-hours question is different from providing medical, financial, or legal advice. Escalation to a person becomes part of the service design once a conversation moves beyond routine administration.

Voice assistants capable of answering business calls are not technically new, so BT’s competitive advantage lies less in inventing the category than in distributing it through existing relationships with companies that already buy communications and connectivity services.

The company is reinforcing that route with an AI Adoption Hub offering guidance and examples alongside the commercial products, effectively treating understanding and deployment as part of the same adoption problem.

As AI spreads beyond early adopters, telecommunications providers, banks, accounting platforms, productivity suites, and other established suppliers may become important distribution channels precisely because they already sit inside routine business operations. BT’s receptionist is a modest example, but its commercial test is correspondingly concrete: whether fewer calls go unanswered and whether the saved staff time is worth the cost of the service.