Aflac's AI Strategy: How the Supplemental Insurance Leader Is Building Conversational Claims

TL;DR#

Aflac is the largest supplemental insurance carrier in the United States, serving more than 50 million policyholders across the US and Japan and generating roughly $18.7 billion in 2024 revenue. Its category-defining promise — "One Day Pay," a claim filed and paid in a single business day — is the clearest example in insurance of customer experience as a competitive moat. Conversational AI insurance tooling extends that moat in two directions: voice-first First Notice of Loss (FNOL) to compress the gap between a covered event and a payout, and continuous policyholder research so Aflac can hear why members file, not just what they file. Aflac's Japanese subsidiary — the country's largest cancer insurer with more than 16 million in-force policies — operates in a market where 24/7 voice contact is table stakes, making it a natural beachhead for voice AI. The Aflac case study sits alongside Lemonade's One Day Pay-style speed thesis as proof that "claims paid faster" is the AI North Star for the entire industry.

What makes supplemental insurance structurally different#

Supplemental insurance covers the out-of-pocket costs that primary health, life, and disability policies leave behind — cash benefits paid directly to the policyholder for events like cancer diagnoses, accidents, hospitalizations, and short-term disability. Unlike a traditional health insurer that reimburses providers, a supplemental carrier writes checks to the member. That structural difference is why "speed of claim" is the entire product. The benefit only matters if it arrives in time to cover the medical bill, the rent, or the missed paycheck.

Aflac, founded in Columbus, Georgia in 1955, built its business around exactly this thesis. Roughly 70% of its revenue still comes from Aflac Japan, where the company has been the #1 cancer insurance provider for more than 40 years (per the Aflac 2024 Annual Report). The remaining ~30% comes from Aflac US — workplace voluntary benefits sold predominantly through payroll deduction at small and mid-sized employers, with the iconic Aflac Duck mascot driving more than 90% unaided brand recall, per LIMRA-tracked awareness studies.

The category itself is large and growing. According to LIMRA's Workforce Benefits research, more than 80% of US workers say workplace voluntary benefits are important to them, and supplemental health products posted record sales in 2024. The structural opportunity for AI is that supplemental claims are high-volume, low-complexity, and emotionally loaded — the policyholder just got hurt, sick, or diagnosed. That combination is exactly where voice-first conversational AI outperforms forms.

The One Day Pay promise as a North Star for AI#

Aflac launched One Day Pay in 2015, committing that eligible claims submitted with complete documentation by 3pm ET would be processed, approved, and paid within one business day. The company has reported processing more than 18 million claims this way. In a recent annual letter, Aflac noted that more than 70% of US individual health claims received online were paid within four days, and roughly half within one — a metric most carriers can't approach.

That promise sets the terms of the AI roadmap. If the brand asset is time to payout, every AI investment has to either:

1. Shorten the time from incident to first notice of loss (FNOL).
2. Shorten the time from FNOL to adjudication.
3. Shorten the time from adjudication to disbursement.

Forms are the bottleneck in step one. A policyholder who just left an emergency room is not going to log into a portal, hunt for a policy number, and complete a 14-field PDF. Voice-first conversational claims intake — the kind powered by the Interviewer agent and similar AI voice systems — collapses that step into a phone call the member can take from a hospital parking lot. The downstream effects compound: faster intake means earlier triage, which means earlier payout.

This is the same speed thesis that animates Lemonade's conversational AI insurance playbook — the fastest-growing AI insurance company in pet insurance, which built its brand on "three seconds to a claim payment" via its AI Jim bot. Lemonade and Aflac sit at opposite ends of the insurance market (a venture-backed insurtech vs. a 70-year-old Fortune 500 carrier), yet both treat speed of claim as the single most important customer-experience metric. That convergence is the signal.

Where Aflac is investing in AI#

Aflac's 2024 annual report and subsequent investor communications describe AI investment across three explicit areas: claims automation, agent-facing productivity (call center and field agents), and underwriting/risk modeling. Like most large carriers — see State Farm's AI roadmap, Liberty Mutual's AI strategy, and Travelers' conversational underwriting shift — Aflac frames AI primarily as a way to scale human capacity rather than replace it. The duck stays.

In practice, that strategy plays out in three workstreams:

Claims automation. OCR and document AI on medical records and Explanation of Benefits forms; rules-based decisioning for clean claims; LLM-assisted summarization for adjusters reviewing complex cases. This is table-stakes work — most top-10 carriers have shipped similar capability — but it's where the "One Day Pay" promise gets defended against rising claim volume.

Agent productivity. Aflac sells through more than 70,000 US producers, the largest workplace-benefits sales force in the country. AI copilots for licensed agents — quote generation, plan-comparison summaries, follow-up drafting — let a 70,000-agent network move with the responsiveness of a much smaller insurtech. This is the same playbook Farmers Insurance is running for its captive agency network.

Underwriting and risk modeling. Particularly relevant in Japan, where Aflac is integrating new cancer insurance products with health-checkup data and AI risk scoring. The challenge here is what every life and supplemental carrier faces — see Prudential's AI life insurance strategy and MetLife's AI strategy for group benefits — moving from 20-question health questionnaires toward conversational underwriting that adapts based on prior answers.

The gap, across all three: Aflac (like most carriers) is investing heavily in AI that processes claims faster, but comparatively little in AI that listens to policyholders at scale. That's the next frontier.

Conversational claims intake: voice-first vs. portal-form#

A voice-first claims intake replaces the FNOL form with a 5-minute conversation the policyholder has with an AI interviewer. The system asks about the incident, captures dates and providers in natural language, follows up on vague answers ("you said you went to the ER — was that a visit or an admission?"), and routes complete files into the adjuster queue. Compared to portal-form FNOL, voice intake produces three measurable advantages:

This isn't theoretical. The convergence of speech-to-text accuracy, LLM-driven dialogue, and SOC 2 / HIPAA-aware infrastructure means the technical pieces are production-ready in 2026. The bigger barriers are integration with claims platforms (Guidewire, Duck Creek, in Aflac's case proprietary systems built in-house) and regulatory comfort with AI-handled intake. Both are solvable — see AI for insurance claims processing in 2026 for the full FNOL transition map.

The framing matters too. Conversational AI in insurance shouldn't be about deflection — keeping the policyholder out of the call center to save cost. For Aflac specifically, the One Day Pay brand cannot be defended by deflection metrics. It has to be defended by speed and clarity of the actual claim experience. Conversational AI is the lever for that.

The Japan angle — supplemental insurance in a voice-IVR-heavy market#

Aflac Japan is the company's largest profit center and operates in a market culturally and structurally suited to voice AI. Japanese consumers, particularly the 65+ segment that represents the bulk of cancer-insurance policyholders, are more comfortable with voice and phone-based service than with self-service web portals. Aflac Japan has historically run extensive IVR (Interactive Voice Response) systems for claims status, premium changes, and beneficiary updates.

That IVR layer is the obvious replacement target for conversational AI. Where IVR forces a caller through menu trees ("press 1 for claims, press 2 for billing"), a modern voice agent listens to a natural-language description of why the policyholder is calling and routes intelligently. For an Aflac Japan policyholder in their seventies who needs to file a cancer-treatment claim, the difference is enormous: instead of navigating a five-level menu, they explain the situation once and the AI captures the FNOL directly.

The Japan market also gives Aflac a privileged data set for AI training. With more than 16 million in-force cancer policies and decades of claims data, Aflac Japan has more cancer-claim conversations recorded than any other organization on earth. That's the kind of vertical-specific corpus that turns generic LLMs into a domain-expert voice agent.

It's worth noting how this differs from how a US health insurer would approach the same problem. Compare it to UnitedHealth's care-navigation AI strategy, Humana's Medicare Advantage focus on senior care navigation, or Cigna's conversational care navigation — all of which use AI primarily to route members through complex care decisions. Aflac's job is structurally simpler: the policy paid out; how fast can we cut the check?

The Lemonade lesson — sister insurtech, same speed thesis#

The most useful comparator for Aflac's AI roadmap isn't another supplemental carrier. It's Lemonade. Despite operating in entirely different segments (pet, renters, and home insurance vs. workplace supplemental health), Lemonade and Aflac converge on a single product philosophy: the claim experience is the brand.

Lemonade's AI Jim claims bot has, at peak, paid claims in under three seconds. It does this by collecting the FNOL through a conversational interface, running anti-fraud heuristics, and approving low-complexity claims with no human in the loop. For deeper context on that strategy, see our full Lemonade case study on conversational AI insurance, which has been one of the highest-trafficked posts in our insurance cluster.

The lesson Aflac can take from Lemonade isn't the tech stack — Lemonade is a digital-native insurtech with no captive sales force. It's the principle: every step between covered event and disbursement is a competitive surface. One Day Pay was already a generation ahead of the industry in 2015. The conversational version of One Day Pay — "describe what happened, and we'll handle it" — is the 2026 update.

For other carriers reading along, the same frame applies. See Geico's AI chatbot strategy, USAA's AI customer service, Allianz's AI customer research, and Chubb's AI strategy for parallel playbooks at carriers with different structural starting points.

How Perspective AI fits#

Perspective AI is the conversational customer-research layer that lets supplemental carriers like Aflac listen to policyholders at the scale and depth that claims data alone can't deliver. Three concrete use cases for an Aflac-shaped business:

Voice-first claims intake research. Before deploying a conversational FNOL bot, you need to know what policyholders actually want it to ask. Perspective AI runs hundreds of conversational interviews with recent claimants — "walk me through what it was like to file your claim" — and surfaces the friction points a survey would never catch. That research output becomes the script for the voice agent.

Policyholder voice-of-customer programs. Aflac, like every large insurer, runs NPS and CSAT programs. NPS scores tell you that satisfaction dropped from 62 to 58. They don't tell you why. Perspective AI's conversational interviews replace the open-text comment box with a dialogue that probes for the underlying reason. The output is what the 2026 voice of customer voice report calls a "VoC program built on voice, not forms."

New product validation. Aflac launches new riders and products regularly (cancer plans in Japan, accident plans in the US, telemedicine bundles). Perspective AI runs concept tests, jobs-to-be-done interviews, and pricing-sensitivity research at scale via the Interviewer agent, letting product teams validate before underwriting.

These use cases all share a common shape: replace the survey or form layer of policyholder research with a conversation. That's the same shift Aflac is making on the claims side. The category is converging.

For carriers ready to start, browse use cases or start a research study to run your first conversational policyholder interview in under an hour.

Frequently Asked Questions#

Is Aflac actively using conversational AI for claims today?#

Aflac uses AI primarily for claims automation and adjuster productivity today, with voice-first conversational FNOL emerging as a 2026 priority. The company's One Day Pay program already leans on document AI and rules-based decisioning to clear clean claims rapidly, and Aflac has signaled investment in agent-facing AI copilots. Full voice-first conversational claims intake — where a policyholder describes the incident to an AI agent that captures the FNOL end-to-end — is still rolling out across the industry and represents the next compression in the form-to-payout cycle.

What is "One Day Pay" and why does it matter for AI strategy?#

One Day Pay is Aflac's promise to pay eligible US claims within one business day of receiving a complete claim. It matters for AI strategy because it sets the customer-experience benchmark every AI investment has to defend: shorten time from incident to FNOL, shorten time from FNOL to adjudication, or shorten time from adjudication to disbursement. Conversational AI is the most leveraged way to compress the first step — replacing the portal-form FNOL with a 5-minute voice conversation.

Why is supplemental insurance a good fit for conversational AI?#

Supplemental insurance is a good fit for conversational AI because claims are high-volume, relatively low-complexity, and emotionally loaded — the policyholder just had a covered event. Forms perform worst exactly in those moments; conversation performs best. Add in that supplemental products pay cash directly to the member (not to providers), and the speed-of-claim metric becomes the entire product experience. That makes conversational FNOL a higher-ROI investment than for a primary health insurer.

How does Aflac Japan differ from Aflac US in its AI roadmap?#

Aflac Japan operates in a more voice-centric market — older policyholders, established IVR infrastructure, and decades of voice-based claims data — making it a natural beachhead for voice AI. The Japanese subsidiary also brings the largest cancer-claims corpus in the world, which is a privileged training data set for vertical-specific conversational AI. Aflac US, by contrast, focuses AI investment on the 70,000-agent producer network and on payroll-deducted workplace benefits, where the customer experience starts at the employer and runs through to claims.

How does Aflac's AI strategy compare to Lemonade's?#

Aflac and Lemonade converge on the same product philosophy — that claim speed is the brand — despite being in different segments. Lemonade pays some claims in under three seconds via its AI Jim bot, treating FNOL as a fully conversational, low-touch experience. Aflac's One Day Pay does this at scale in the supplemental category, with more incumbent infrastructure to upgrade. The shared lesson is that conversational AI is most valuable when it compresses the gap between covered event and disbursement — not when it deflects calls or replaces humans.

What should a supplemental insurance carrier do first to deploy conversational AI?#

A supplemental carrier's first move should be running conversational policyholder research to understand the actual claims experience before deploying any customer-facing AI. Talk to 100+ recent claimants in conversational interviews; identify the friction points, the moments of confusion, and the language they use. That research becomes the foundation for both the voice agent script and the broader claims-experience redesign. Most carriers skip this step and ship AI on top of assumptions; the higher-ROI path starts with listening.

Conclusion#

Aflac's One Day Pay promise has been the highest-leverage customer-experience commitment in supplemental insurance for a decade. Conversational AI insurance tooling is what defends and extends that promise for the next decade — voice-first FNOL that turns claims intake into a 5-minute conversation, conversational policyholder research that replaces survey-driven VoC with dialogue, and a Japan-market beachhead where voice AI fits the policyholder behavior already in place. The carriers that win the 2026–2030 window will be the ones that treat every minute between covered event and disbursement as a competitive surface — and Aflac, with its existing speed brand and 50M-policyholder data set, is uniquely positioned to win that race.

Perspective AI is the conversational research layer for carriers running this play. To see how supplemental and primary insurers are replacing survey-driven policyholder research with conversational interviews at scale, start a research study, browse use cases, or compare alternatives to your existing VoC tool.