An AI call center for utilities has AI voice agents answer inbound calls and place outbound ones for an electric, gas, or water provider, and people take the calls that involve safety, hardship, or a decision. Bland builds AI voice agents for 250+ enterprise customers, and Bland's pricing page lists what each plan includes.
Inbound agents take outage reports, balance questions, and start-service or stop-service requests, and outbound agents place due-date reminders and planned-work notices.
A utility does not hand its phone lines to AI in one move: a team picks one queue, keeps a person behind every agent, and widens the share as its numbers hold. The wider case for large contact centers sits in Bland's enterprise guide.
What is an AI call center for a utility?
An AI call center for a utility is a phone operation in which AI voice agents handle inbound and outbound calls for an electric, gas, or water provider, and people handle the calls an agent should not.
An inbound agent answers a utility's number, works out why a customer rang, reads the answer from a connected system, and either finishes the call or transfers it. An outbound agent places a call, such as a due-date reminder, and works the same way from the other direction.
Volume at a utility arrives in waves, so a storm or a billing run can multiply calls in a day. Many callers want one fact, such as when power returns or when a bill is due, and a voice agent takes a full sentence where a phone tree takes four keypresses. Bland's calls run at sub-200ms latency, which matters on a phone line because a caller waits through every pause.
Two parts of the job weigh more at a utility than at most businesses.
Safety comes first, because a caller who smells gas or reports a downed wire needs a person and emergency services before a ticket number.
Fairness comes second, because a customer behind on a bill may be in real difficulty, and a handoff to a person should be quick and should carry the details so the caller does not start over.
Which utility calls can an AI voice agent take, and which stay with people?
An AI voice agent suits utility calls that repeat, follow fixed steps, and end with an answer from a system, and people keep the calls that involve safety, hardship, or a judgment only a person has the authority to make. Inbound, the calls that fit an agent usually look like these:
- Outage reports, and status checks on whether power is back
- Balance, due-date, and payment-status questions
- Start-service, stop-service, and move requests
- Questions about a bill line or a rate plan, answered from an approved knowledge source
Outbound, the work is notices and reminders:
- Payment due and payment missed reminders
- Planned-work notices, such as a scheduled interruption on a street
- Weather and restoration updates for customers who asked to receive them
Status questions fit best because the answer lives in a system. The catch is the connection, because an agent that cannot read a utility's systems is a polite menu.
What stays with people is shorter and harder to automate.
A caller who smells gas, reports a downed line, or describes sparking equipment is told how to reach emergency services and gets a person at once. A customer in hardship, a customer who depends on powered medical equipment, a disputed bill, a disconnection notice, a complaint, a bereavement, and any request needing authority an agent was never given all reach a person too.
For those calls an agent's job is to hand over well, and on Bland a warm transfer is an Enterprise feature that Start and Build don't include.
Can an AI voice agent handle outage and storm calls?
Yes, an AI voice agent can take outage reports and restoration questions, and a person takes any caller who reports a danger. Outage calls are repetitive, structured, and urgent, so an agent can collect:
- The service address or account number
- What the caller sees: lights out on one street, a flicker, or a partial loss
- Whether anyone at the address depends on powered medical equipment
- Whether anything dangerous is in view, such as a wire on the ground
A utility decides whether an agent asks about medical equipment, where the answer is kept, and what happens next. An agent that hears yes hands the call to a person and never promises that power returns sooner.
The agent writes a summary into a utility's outage system and tells the caller what that system says. A restoration estimate is a figure someone else produced, so an agent reads it out as an estimate, and when none exists it says so.
A storm sends calls in a surge, and agents take outage reports in parallel while dispatch works the ones that matter most. On Enterprise, concurrency is contracted to a utility's volume, so a utility that expects storm peaks should raise them with Bland's sales team.
Write a safety rule before launch. A utility's safety leaders decide which words move a call straight to a person, such as gas, sparking, wire down, or fire, what an agent says about emergency services before it hands over, and whether it tells a caller to leave a building before anything else.
An agent follows that list from the first second of a call. Test it with callers who bury the word mid-sentence.
Can an AI voice agent take billing and payment calls for a utility?
Yes, an AI voice agent can answer billing and payment-status calls for a utility, and people take payment arrangements, disputes, and hardship. A caller who asks what they owe or whether last week's payment posted is asking a question whose answer sits in a billing system. Before an agent reads out a balance or a payment, it confirms who is calling under a utility's own identity rules.
An agent can also explain a bill line or a rate plan from an approved knowledge source. A knowledge source can be a file, a site, a block of text, or a synced document, and an agent it is attached to searches it during the call.
Outbound, an agent can place due-date and missed-payment reminders to customers who have consented to them, and never as cold calls. Keep a reminder to a date, an amount, and a way to reach a person, and hand the call to a person the moment it turns to hardship, a dispute, or a payment arrangement.
A missed-payment reminder is also a call about a past-due amount, so collection rules can reach it, and a utility's counsel decides what an agent may say on one. Keep reminders separate from any notice a utility must give before it shuts off service, and let a utility's counsel decide what a reminder may say about disconnection.
A team settles these points before launch:
- How an agent names the utility and says an AI voice agent is calling, at the start of each call
- Where a customer's consent is recorded, and how an agent checks it before it dials
- Which days and hours an agent may call, and how often it may call one customer
- What happens when a customer says to stop calling
- What a voicemail may say about a balance, and what an agent does when someone other than the account holder answers
- Which words move a call to a person, such as can't pay, dispute, medical, or shutoff
Can an AI voice agent handle start, stop, and move requests?
Yes, an AI voice agent can take start-service, stop-service, and move requests, and a person takes the exceptions. An agent confirms who is calling under a utility's own identity rules, collects the service address and the date, and writes the request into a customer system.
A person should take cases that break the pattern, such as a request that changes who holds an account, an address a system cannot find, or a customer with an unpaid balance who asks to move service.
Which rules apply to an AI call center for utilities, and what does Bland hold?
Rules that can apply to an AI call center for utilities include the TCPA's consent rules for calls made with an artificial voice, and a utility's counsel decides how they apply to a given call. That is a description of a rule and not legal advice.
The TCPA requires the prior express consent of the called party for a call made with an artificial voice, absent an emergency purpose or an exemption. The TCPA's rules define emergency purposes as calls made necessary in any situation affecting the health and safety of consumers. The same rules require an artificial-voice message to state clearly, at the beginning, the identity of the business responsible for the call.
Whether a particular outage notice, safety alert, or payment reminder falls inside that definition is a question for a utility's counsel, and an outbound program should hold a record of consent either way.
Under a federal rule in effect from 25 November 2026, a caller can designate an exclusive way to revoke consent: an automated voice or key-press opt-out on the call, set reply words, such as stop, in a text, or a designated website or phone number.
Exclusivity depends on a caller disclosing the designated method clearly and conspicuously on the call or in the text, and a caller that designates none must honor revocation made by any reasonable means.
Either way, a revocation must be honored within a reasonable time and no later than ten business days, and a revocation made in reply to an informational call may be read as covering only the kind of calls it was aimed at, and counsel decides how far a given request reaches.
A utility decides with counsel whether its calls designate a revocation method, and which one, and discloses it on the call. An outbound program records every customer request to stop calling, whichever method it designates.
A designation does not change the automated voice or key-press opt-out the rules already require on an artificial-voice call made under an exemption from consent.
On Bland's side, the compliance list reads: SOC 2 Type II, HIPAA compliant with a BAA, GDPR compliant with a DPA, and PCI DSS v4.0.1 self-attested. A BAA, SSO, and data residency come with Enterprise. On FedRAMP, the record reads: FedRAMP 20x Class C (Moderate), certified 17 November 2026, FedRAMP Marketplace FR2628647242.
PCI DSS is the card industry's standard for card data. Bland's PCI DSS v4.0.1 status is self-attested, so a utility that wants to take card payments on a call should ask Bland how card data is handled and decide whether that status meets its own program.
A security review counts the companies a call passes through. Bland self-hosts its models and uses no third-party model providers. Bland operates a closed-source model it developed itself. With no outside AI lab in the call path, a security team reviews one AI company and not several.
Bland's Trust Portal holds its SOC 2 Type II report, a penetration test summary, its Data Processing Agreement, and its Business Associate Agreement, and security teams can file a security review request or ask for customer references. Bland runs quarterly third-party penetration tests. Bland's trust and security page lists the program in full.
Bland's compliance status does not make a utility's call flow compliant with any rule. A caller should also hear at the start that an AI voice agent is on the line, and how to reach a person.
How does an AI call center connect to a utility's outage and billing systems?
An AI call center connects to a utility's customer information, billing, and outage systems through Bland's REST API and webhooks. Salesforce, HubSpot, Cal.com, Calendly, Notion, and Slack connect from Bland's dashboard. Anything else connects through Bland's REST API and webhooks, which Bland's forward-deployed engineers can help with on Enterprise.
Bland's integrations list what connects from Bland's dashboard.
Two decisions come first: which system holds the record an agent reads, such as a balance, an outage status, or a service address, and which system an agent writes to when the call ends, such as an outage report, a service request, or a callback task. A system connects only if it exposes an API, so a team confirms that before it plans a launch.
Plan the read before the write. An agent that reads a due date is a small first launch, and one that opens outage tickets needs reviewed required fields and a person watching the first batch.
Every agent has development, staging, and production environments, and live calls run the pinned production version. Evaluations score an agent's calls, and a passing run can be required before a version is promoted.
What results has Bland produced, and what do they say about a utility?
Idaho Housing and Finance Association, a housing finance association and not a utility, shows what Bland did on one high-volume inbound line, and no utility customer is named on this page.
Idaho Housing and Finance Association saves $750,000 a year by replacing its IVR with Bland, which handles 4,000 inbound calls a day and cut average call time by 20%.
"We save thousands of dollars a day."
Greg Blake, Chief Information Officer, Idaho Housing and Finance Association
Those figures are one organization's own report and not a forecast, so a utility should size its own queue before it borrows a number.
What does a utility call look like on an AI voice agent?
An outage call on an AI voice agent runs a few turns: it says it is an AI, collects the address, reads the status from a connected system, and moves to a person the moment a danger comes up.
Sample call (illustrative). This exchange is written for this page and is not a recording of a real call or customer, and a utility's own safety and identity rules would set its real steps.
- Agent: Thanks for calling about a power problem. I'm an AI voice agent, and I can bring in a person at any point. What's the address where the power is out?
- Caller: 218 Alder Street.
- Agent: Thank you. I see an outage reported on Alder Street, and the estimate I have is that power returns by 6 this evening. That is an estimate and it can change. Is anyone at the address on medical equipment that needs power?
- Caller: No, all fine here, but there's a branch across the road and a wire is hanging down.
- Agent: A downed or hanging wire can be dangerous, so please keep well away from it and keep others away too. If anyone is in danger right now, hang up and call 911. If not, stay on the line and I'll connect you to a person with everything you've told me.
The agent read the estimate as an estimate and made no judgment about the wire. A utility's safety leaders wrote the words an agent says when a caller reports one, and the agent said them and handed over with the details attached.
How does a utility add AI voice agents without replacing its contact center?
A utility adds AI voice agents by putting them in front of or beside the contact center it already runs, and routing what an agent cannot finish to people inside that platform. Bland is an accredited Five9 AI Agent Connect partner and a NICE DEVone technology partner.
In front means an agent answers first. Beside means an agent takes one queue, such as after-hours calls or storm overflow, while the contact center platform keeps routing the rest.
Bland's SIP connection is built and tested once in Five9's partner environment and reused for each customer. Calls go from Five9 to a Bland agent and transfer back to Five9 skills or agents, with call data in SIP headers. NICE built and tested the Bland trunk once and adds each customer Business Unit to it, with US and EU connections. Queues, routing, and agent desktops stay where they are.
Every enterprise customer runs on its own isolated infrastructure, run by Bland: dedicated compute, containers, and resources, with releases the customer chooses when to adopt and canary deployments on a share of real calls.
Dedicated GPUs run by Bland are available. Bland publishes a 99.9% uptime SLA, and a utility whose customers phone during storms should ask for the terms in writing, including how uptime is measured and what a customer receives when the measure is missed.
A utility's team can build and edit agents in plain English with Norm or Bland's visual builder, without writing code. Connecting a utility's own systems through Bland's REST API and webhooks may need engineering work on a utility's side, and Bland's forward-deployed engineers help with that on Enterprise.
A rollout runs in stages:
- Pick one queue and one call type, such as outage status, with read-only access to one system.
- Send a share of those calls to an agent and transfer the rest to people.
- Read a sample of calls every week, and widen the share only when transfer rate, call length, and hang-ups hold.
The same line between agent and person applies in other industries. Bland's telecom guide covers service and billing calls, Bland's banking guide covers payment calls, and Bland's healthcare guide covers patient calls.
How much does an AI call center for a utility cost?
An AI call center for a utility pays for the minutes an AI voice agent spends on calls, and Bland's headline rate is $0.14 a minute on the self-serve Start plan, or $0.12 a minute on self-serve Build with a $299 monthly platform fee. Enterprise pricing is custom, set with Bland's sales team.
Phone is billed separately, on a customer's own carrier or Bland's at pass-through cost.
Three assumptions carry the arithmetic, and a utility should replace each with its own figure. Assume a service line takes 60,000 calls a month, assume AI voice agents take 35 percent of those calls, and assume an average call lasts 4 minutes.
Thirty-five percent of 60,000 is 21,000 calls, or 84,000 minutes a month.
At self-serve Build's $0.12 a minute, that is 10,080 dollars plus the $299 monthly platform fee, 10,379 dollars a month before phone. At self-serve Start's $0.14 a minute, the same minutes come to 11,760 dollars, and Build is the cheaper plan above 14,950 minutes a month.
Start also allows 100 calls a day, and 21,000 calls a month is about 700 a day, so this volume belongs on Build or Enterprise.
A 4-minute call costs 48 cents on self-serve Build and 56 cents on self-serve Start, before phone.
A utility that needs warm transfers to a person, SSO, or data residency is on Enterprise from the first day, where concurrency is contracted to its volume and pricing is set with Bland's sales team, so the self-serve rates here are a reference point and not a quote.
Bring a fully loaded cost per handled minute for the calls an agent would take to that conversation. Bland's pricing page lists every plan row.
Frequently asked questions
Can AI replace a utility's call center agents?
Not all at once. AI voice agents take repeat calls such as outage status, balance questions, and reminders, and people keep safety calls, hardship, and disputes. Bland's Enterprise plan includes warm transfers to a person.
What happens when a caller reports a gas smell or a downed line?
An agent follows a safety rule that a utility's safety leaders write before launch: it sets what a caller hears first about emergency services and about leaving a building, and moves the call to a person at once. A warm transfer is an Enterprise feature.
Does Bland connect to a utility's billing and outage systems?
Yes, through Bland's REST API and webhooks, which can need engineering work on a utility's side. Bland's forward-deployed engineers help with that on Enterprise.
Can Bland place outbound outage and payment calls?
Yes, such as payment reminders and planned-work notices, to people who have consented to them. A utility's counsel decides how the TCPA applies.
What does Bland cost for a utility?
Self-serve Start costs $0.14 a minute, and self-serve Build costs $0.12 a minute plus a $299 monthly platform fee. Enterprise pricing is custom, set with Bland's sales team. Phone is billed separately, on a customer's own carrier or Bland's at pass-through cost.
