AI startup “WagerGPT” and more chase $150B sports gambling market with mixed results

A growing number of entrepreneurs are developing AI-powered gambling tools, betting that artificial intelligence can give ordinary bettors an edge in the rapidly expanding online sports betting market. While some claim success rates of 56-60 percent compared to standard odds of 52 percent, the field remains largely experimental with no clear wave of newly minted millionaires emerging from AI-assisted gambling.

The big picture: The convergence of America’s booming $150 billion sports betting industry with the AI gold rush has created a cottage industry of startups promising to turn bots into winning bet-makers.

• Americans spent over $150 billion on sports-related wagers last year, with the American Gaming Association reporting a nearly 24 percent jump in sports betting popularity in 2024.
• Federal legalization of sports betting seven years ago coincided with the modern AI boom, creating opportunities for tech-savvy entrepreneurs to combine both trends.

Key players and their approaches: Several startups are taking different strategies to merge AI with gambling, from tip services to fully automated betting agents.

• Carson Szeder’s MonsterBet charges $77 monthly for MonsterGPT, which uses projection models and web scraping through retrieval-augmented generation to select bets on professional sports.
• Massachusetts-based Rithmm offers “AI-powered sports intelligence” packages starting at $30, while JuiceReel provides a free basic app for similar purposes.
• Siraj Raval’s WagerGPT charges $199 monthly and claims to place bets automatically, though user feedback suggests the service is “completely dead” and a “waste of money.”

In plain English: Retrieval-augmented generation (RAG) allows AI tools to pull in fresh information from the internet on top of what they learned during training—like giving a calculator access to current sports statistics instead of relying only on old data.

Technical challenges: Companies attempting fully automated betting face significant hurdles in actually placing wagers rather than just providing tips.

• Tom Fleetham’s experience with blockchain platform Zilliqa’s horse racing AI agent Ava highlighted timing issues: “Where it got hard was actually trying to place the bets… It took forever. We gave up.”
• Most automated betting products focus on cryptocurrency-based platforms since AI agents cannot control traditional bank accounts but can operate crypto wallets.
• FanDuel’s approach with chatbot Ace deliberately avoids automation: “The power to place bets should always stay in the hands of our customers themselves—not bots, AI agents, or anything (or anyone) else,” says VP Jon Sadow.

Crypto-gambling fusion: Several projects are attempting to create AI-powered betting hedge funds using cryptocurrency infrastructure.

• Sire (formerly DraiftKing) operates as a decentralized autonomous organization that plans to pool customer funds, place bets on crypto-based sports books like Polymarket, and redistribute winnings minus performance fees.
• Max Sebti’s Score company uses “computer vision tech that is watching the games” combined with public and private data to determine winning bets.
• Coinbase’s AgentKit enables AI agents to execute various financial transactions including sports betting, though product manager Lincoln Murr admits: “How profitable these agents truly are, I don’t know.”

In plain English: A decentralized autonomous organization (DAO) is essentially a member-owned investment club that uses blockchain contracts instead of traditional paperwork—like a betting pool where the rules are coded into the system rather than managed by a single company.

Marketing pivot: Some companies have shifted from actual betting automation to creating promotional AI agents for established gambling platforms.

• QStarLabs’ Memetica “fully pivoted” from crypto token promotion to designing AI agents for over a dozen iGaming clients.
• Their agent Divinia posts promotional content on X for decentralized platform Divvy, essentially serving as a marketing gimmick rather than a functional betting tool.

What they’re saying: Industry participants acknowledge both the potential and the persistent challenges in AI gambling.

• “We have some people who are probably hitting around 56 to 60 percent of the time, myself included,” claims MonsterBet’s Szeder about his AI tools’ success rate.
• Yang Tang from QStarLabs warns about traditional scams being repackaged with AI branding: “We’re not nefarious people, so we don’t do that with our agents. But all of these tricks are well known.”

Why this matters: These early experiments represent how AI agent hype often outpaces actual utility, while highlighting the beginning stages of merging gambling and artificial intelligence technologies that could reshape both industries.