Almost every gambling company now uses AI, but very few run it well. In April 2026, researchers at the UNLV International Gaming Institute worked with KPMG to survey 83 gambling companies and 113 regulators.
More than 80% already use generative AI, they found, yet the industry scored only 45 out of 100 on AI maturity and a worrying 30 out of 100 on governance.
Cevro is the AI player support platform built only for iGaming, with agents that handle questions about bonuses, KYC, payments and responsible gaming from start to finish.
That gap between using AI and running it properly is where the real work sits. Below are 12 ways artificial intelligence is reshaping the industry.
Key Takeaways
Over 80% of gambling firms use generative AI, but governance scores just 30/100.
Fraud tools lead adoption at roughly 85%, followed by personalization at 68%.
Personalized bonuses raise average player spend by about 34% over blanket offers.
Manual responsible gaming checks cover only 2-7% of chats. AI covers all of them.
Cevro agents resolve up to 90% of player questions, with CSAT and NPS at 4.8.
AI Spending in iGaming Reaches $2.9 Billion in 2026
Online gambling took in roughly $115 billion during 2026, up about 12% on the year before. Around 68% now comes from regulated markets, which means more rules and more audits.
Track360 estimates operators will spend about $2.9 billion on AI this year, with fraud and payments tools taking the biggest slice at roughly $870 million, then CRM and personalization at $780 million and support automation at $440 million.
Adoption is uneven, and that tells you something:
Fraud detection sits highest at about 85%, because the cost of getting it wrong is immediate.
Personalization follows at 68%, up from 42% in 2021.
Support automation trails at roughly 60%, but it is growing fastest.
12 Ways AI Is Changing the iGaming Industry
1. AI Support Agents Resolve Player Issues Instead of Explaining Them
For years, chatbots were search engines with a friendly face. A player asked where their money was, and the bot pointed at a help article, but that approach only automated 10-20% of questions.
Modern AI agents work differently because they can act. Take a missing deposit for example, rather than quoting policy, the agent checks identity, pulls the payment gateway logs and reviews KYC status before replying with something useful, such as "Your €50 deposit is pending PSP approval, expected in 15 minutes."
Up to 90% of questions get resolved this way.
2. Responsible Gaming Monitoring Covers Every Player Conversation
When compliance teams review chats by hand, they typically see 2-7% of them. That was fine when regulators only asked whether a policy existed. It stopped being fine once the Malta Gaming Authority began requiring operators to watch deposit frequency, and the UK Gambling Commission turned dozens of harm markers into guidance.
AI reads all of it instead of a sample, and Cevro monitors every conversation for signs such as distress language or deposit chasing before handing the case to a person.
3. KYC Becomes a Continuous Check Instead of a One-Time Gate
Verification used to happen once, at sign-up, but then deepfake attacks rose sharply and a one-time check began to look like a weak point. This matters for support because KYC creates tickets at the worst moment, right when a new player is deciding whether to stay.
Cevro's work with Sumsub lets agents handle document uploads, rejections and age checks inside the chat itself.
4. Fraud Detection Maps Networks Instead of Single Accounts
Older systems checked each account on its own, so a well-built fake looked normal. Newer ones map connections instead, such as shared devices, linked cards and overlapping IP addresses. One fake account looks clean, but the ring behind it does not.
Some systems report up to 95% accuracy, yet 77% of fraud teams still say attacks are evolving faster than their defences.
5. Bonus Spend Gets Targeted at the Right Players
Bonus spend is one of the biggest costs an operator controls, so it is an obvious place for a model. Personalized offers lift average player spend by roughly 34% over sending everyone the same thing.
There is a catch: Clever bonus rules create confusing disputes, and those land in support.
Cevro trains agents on the full promotional rulebook, so use cases such as missing free spins get sorted without a handoff.
6. Sportsbook Odds Are Priced by Models, Not Traders
Trading was the first part of iGaming to lean on AI and is still the most advanced. More than 80% of major operators run pricing models live, repricing markets in under a second and cutting risk exposure by an estimated 15-25%.
The knock-on effect is volume, because thousands of small in-play markets mean thousands more questions about settlements and disputed odds. Cevro's iGaming solution is built for that load.
7. Churn Prediction Flags Players Before They Leave
Retention teams used to react when a player went quiet, a win-back email went out, and the value was already gone. Prediction models moved that forward, reaching around 81% accuracy at 30 days versus 57% for simple rules.
Support chats are among the best early warnings, yet most operators never read them. Cevro Insights flags frustrated players and alerts teams when a VIP is at risk. As Julija Kozireva, Head of VIP at Alpha Affiliates, put it:
"After analyzing chats with and without Cevro, we've already seen at least a 10% increase in player LTV."
Read Cevro's guide to player retention strategies for more information.
8. Casino Lobbies Personalize Games for Each Player
Game recommendations lift engagement by roughly 30%, which is why personalized lobbies are standard at bigger brands. The interesting part is what happens next.
A player gets a tailored homepage, then opens a chat and meets a script. Cevro's agents read tone and VIP status, so the experience holds up on both sides of the visit.
9. Payment Questions Get Resolved Inside the Back Office
Payment tickets carry the highest risk of escalation, because missing money damages trust fast. As Cevro's glossary puts it, without back-office access an AI can only explain, but with it the AI can act.
That access comes from Cevro's integration layer, which connects to EveryMatrix, SoftSwiss and Playtech, plus tools like Comm100 and Zendesk. In one deployment, escalations fell 40%.
10. Support Conversations Become Usable Operator Data
Most operators can tell you their CSAT score dropped, but very few can say why, because the reason sits in thousands of transcripts nobody has time to read.
Cevro Insights groups those chats into topics automatically, so instead of asking how many tickets arrived this week, teams can ask which bonus is causing complaints right now.
11. New Markets Open Without New Support Hires
Opening a market used to mean hiring for the language, the shift pattern and the local rules. AI breaks that link, but only if it understands each jurisdiction rather than simply translating.
Cevro runs across 100+ languages and 50+ markets. The proof came when Alpha Affiliates launched two brands at once and picked up 6,300 extra chats in a month, and they hired nobody.
The same math applies for BPOs and managed service providers.
12. AI Assistants Change How Players Discover iGaming Brands
Gartner predicted in 2024 that search traffic would fall 25% by 2026 as assistants absorb queries. Track360 now estimates 8-12% of gambling searches start with an AI assistant, up from under 2% in 2023. Because these tools quote data pages rather than adverts, being cited has quietly become a growth channel.
What AI Player Support Costs Compared to Human Agents
Support is expensive, and the headline wage hides most of it. In the United States, iGaming support pays around $19.26 an hour, but once you add benefits, software, training and turnover, one agent really costs closer to $64,500 a year, and tickets average $8 to $12 globally.
Cevro's build versus buy analysis puts a three-year in-house project in seven-figure territory, with 12 to 24 months before it returns anything. Cevro does not publish prices, so the ROI calculator is the practical starting point.
iPlay Gaming Platform reached 91% automation and went from 30 support agents to 15 in stages, while SweepNext Casino passed 90% and kept CSAT above 4.9.
Why AI Governance Decides Whether iGaming Deployments Succeed
That 30 out of 100 governance score is the real story in the UNLV research, only about one in five companies has anyone formally responsible for AI. In most industries a mistake is annoying, but in a regulated market it becomes a reportable event.
Cevro's guardrail framework handles this in three stages:
Before acting, every procedure checks conditions such as authentication and responsible gaming flags, and blocks risky moves outright, so there are no KYC overrides or withdrawal approvals without a human.
During the chat, local rules decide what the agent can say.
Afterwards, every action leaves an audit log.
For enterprise operators, Cevro is SOC2 Type II audited, with PII masking and zero data retention.
The rollout works best in stages as well. Month one usually lands at 40-60% automation, rising to 80-90% on well-defined workflows, and operators handling 100,000+ monthly chats have gone from integration to live in under 30 days.
Keeping procedures current is its own job, since promotions change constantly, which is exactly what CevroScribe is built to handle.
Where AI Still Needs Human Support Agents
It would be dishonest to pretend AI fixes everything, because small operators with simple questions do not need an autonomous agent. Voice is still on the roadmap rather than live, so treat any vendor promising full voice parity with caution.
Humans also stay in the loop because responsible gaming cases and serious complaints go to people immediately, with full context attached.
CEO Chaim Heber describes what happens to the team:
"A lot of agents get upskilled and they start becoming managers and decision makers instead of actually having to communicate back and forth with the players."
Conclusion
Cevro AI is the player support platform built only for iGaming, with agents that open the back office, check the account and fix the problem rather than pointing at a help page.
AI now touches almost every part of the industry, from odds pricing and fraud detection to bonusing and retention.
What separates operators seeing real returns from those still testing is not access to models, but the depth of integration, guardrails that survive an audit, and procedures that keep pace with the business.
To see how that would work with your own tickets and your own stack, book a demo with the Cevro team.
Read Next:
iGaming 'Post-Bonus' Era: Why CX Is the Retention Engine in 2026
The Most Effective Strategies for Player Retention in iGaming
FAQs:
1. What is AI in iGaming?
AI in iGaming is the use of machine learning and language models across operator functions such as odds pricing, fraud detection, KYC checks, bonus personalisation and player support. More than 80% of gambling companies now use generative AI, according to UNLV and KPMG research published in April 2026.
2. How is artificial intelligence changing iGaming customer support?
Artificial intelligence is changing iGaming customer support by resolving issues instead of deflecting them. Older chatbots handled only 10-20% of questions because they could explain policy but never see an account. Agents connected to the back office check payments and verify KYC, which is how Cevro reaches 90% automation.
3. What percentage of iGaming operators use AI in 2026?
The percentage of iGaming operators using AI in 2026 is above 80% for at least one live system. By area, roughly 85% use it for fraud detection, 68% for personalization and 60% for support automation, per Track360's July 2026 review of industry surveys.
4. Is AI safe to use in regulated iGaming markets?
AI is safe to use in regulated iGaming markets when guardrails are built in rather than switched on later. That means checking conditions before every action, blocking high-risk moves such as KYC overrides, keeping audit logs and escalating responsible gaming signals automatically. Governance remains the weak spot at 30 out of 100.
5. How much does AI cost compared to human support in iGaming?
AI costs far less than human support once volume grows. A fully loaded support agent runs about $64,500 a year in the US, with tickets costing $8 to $12 globally. Cevro does not publish pricing, but deployments report a 3x reduction in support costs, while building in-house reaches seven figures over three years.
