AI Finds Decade-Old XRP Ledger Bug, Patched in Days

Key Takeaways

• AI discovered a decade-old XRP Ledger bug that could have minted 18.45 trillion XRP, patched within three days after a $250,000 bounty was paid to Veria Labs.
• 59% of organizations using AI agents lack full integration with identity and access management policies, per a survey of 250 IT leaders.
• Microsoft 365 users show weaker auditability (21% in lowest tier) compared to Google Workspace users (11% in lowest tier).
• The FCC is considering allowing AI-powered political robocalls before the November midterms, with a potential ruling by late October.
• AI founders in Silicon Valley broadly support the Trump administration's self-regulation approach to AI safety.
• A report warns AI, nuclear war, and bioweapons could contribute to human extinction by 2100, identifying 17 major global threats.
• The sci-fi series *The Arc* was produced for $38,000 in AI credits, demonstrating low-cost AI-generated content creation.
• Retail investors at the Silvia Investor Summit identified agentic AI as the top trend they are most excited about.
• Cyderes partnered with Palo Alto Networks to integrate Cortex XDR and Cortex XSIAM into managed security operations across the U.S., Canada, and U.K./EMEA.
• Bulk purchases of Korean used books by a Arizona distribution firm raised suspicions of AI model training data collection.

AI finds XRP Ledger bug that could have created 18.45 trillion XRP

Veria Labs discovered a flaw in the XRP Ledger that could have minted 18.45 trillion XRP, far exceeding the 100 billion coin supply. Researcher Cayden Liao reported the bug on September 22, 2026, and engineers patched it within three days. Over 80% of validators installed the fix immediately. RippleX found no evidence the bug was exploited.

AI uncovers XRP Ledger flaw threatening $94 billion market

An AI security agent found a decade-old flaw in the XRP Ledger that could generate 18 trillion XRP in one transaction. The bug involved integer overflow errors in the payment engine and supply protection system. Veria received a $250,000 bounty for the discovery. Developers bypassed normal governance rules to deploy an emergency fix on servers running version 3.4.1.

AI exposes hidden XRP bug missed by human reviewers for ten years

An AI security system found a flaw in the XRP Ledger that human reviewers missed for a decade. The bug let sellers receive full payment while buyers were charged nearly nothing, creating fake XRP. Engineers fixed it within three days after it was reported. AI has also uncovered Bitcoin vulnerabilities since July, including flaws in the Coldcard wallet and Core Lightning software.

FCC weighs allowing AI-powered political robocalls before midterms

The FCC is considering a petition from the Club for Growth to let political groups use AI or prerecorded voices in unsolicited phone calls. Consumer groups warn this could flood voters with misleading automated calls before the November midterms. The Republican-led FCC advanced the request for public comment and could rule by late October. Anna Gomez, the sole Democrat commissioner, opposes the move.

FCC may permit AI robocalls targeting voters ahead of midterms

A petition asks the FCC to allow political robocalls using artificial voices on cellphones without prior consent. The Club for Growth argues current rules are outdated and limit free speech. Tim Harper from the Center for Democracy and Technology warns the change could help campaigns deceive voters. Past deceptive robocalls have led to criminal charges and fines, including a $15,000 fine against Jacob Wohl and Jack Burkman.

Cyderes Partners with Palo Alto Networks to Boost AI Security

Cyderes announced a partnership with Palo Alto Networks to combine its managed detection and response services with Cortex XDR, Cortex XSIAM and Idira. The goal is to modernize AI driven security operations for enterprises in the US, Canada and U.K./EMEA. This deal makes Palo Alto Networks' AI SecOps platforms the technology backbone for Cyderes' global SOC offering. It could strengthen demand for higher margin recurring security subscriptions across endpoint, cloud and identity.

AI Founders in Silicon Valley Back Self Regulation on Safety

AI founders in Silicon Valley cheered calls from the Trump administration for the industry to police itself on safety. The movement gained support among tech leaders who want the industry to set its own rules. The article highlights the growing debate over how AI companies handle safety concerns. This self regulation approach could shape future AI development practices.

AI Sci Fi Series The Arc Made With 38,000 Dollars in Credits

Filmmakers created the AI generated sci fi series The Arc using 38,000 dollars in AI credits. Each shot required 8 to 10 generations to produce. The creators used tools including Seedance and Artlist to complete the project. This series shows how AI technology is changing the way films are made. It demonstrates that AI can produce full length content at a low cost.

Retail Investors Are Bullish on Agentic AI at Summit

The author attended the Silvia Investor Summit in New York City where retail traders showed strong excitement about agentic AI. Attendees included day traders, crypto enthusiasts and market junkies who agreed on one major trend. Most investors said agentic AI is the game changer they are most focused on right now. The crowd was mega bullish about how agentic AI could reshape finance and markets.

AI, Nuclear War and Bioweapons Could End Humanity by 2100

A report warned that AI, nuclear war and bioweapons may trigger the end of humanity by 2100. The commission identified 17 major threats including climate change, antimicrobial resistance and air pollution. The report used disability adjusted life years to measure the impact of each threat. It found that limiting global warming to 1.8 degrees Celsius could prevent the loss of 1.89 billion healthy life years. The researchers plan to update their findings every year.

AI Agent Security Gaps as Integration Lags Behind Adoption

A survey of 250 IT leaders in the U.S. and U.K. found that 59% of organizations using AI agents lack full integration with identity and access management policies. JumpCloud introduced an Agentic IAM Lifecycle framework to help manage and govern AI agents with clear access controls and auditing. The study also revealed that 21% of Microsoft 365 users fall into the lowest auditability tier compared to 11% of Google Workspace users. These gaps present new security challenges and service opportunities for managed service providers.

Bulk Used Book Purchases in Korea Spark AI Training Fears

A bookstore in Seoul received an email from a book distribution firm in Arizona wanting to buy large quantities of Korean used books for resale. The request excluded children's books and offered advance payments, raising suspicions that the books were meant for AI model training. Similar bulk purchase attempts were reported in the United States, Germany, Ireland, and Japan. Experts believe AI companies are buying used books for training before eventually destroying them.

Sources

• AI Uncovers XRP Ledger Bug: Is It a Threat or a Defense?
• AI discovers XRP Ledger flaw that could mint 18 trillion tokens and put $94 billion market at risk
• AI Uncovers XRP Ledger Bug: Is It a Threat or a Defense?
• FCC considers allowing more political robocalls before midterms, including those created with AI
• FCC considers allowing more political robocalls before midterms, including those created with AI
• Cyderes Partnership Has Investors Rechecking The Bull Case For Palo Alto Networks Stock
• In Silicon Valley, AI founders cheer Trump’s calls for the industry to police itself on safety
• Filmmakers made the AI sci-fi series 'The Arc' with $38k in credits
• I attended a retail investor summit, and the day-trader crowd is mega-bullish on agentic AI
• AI, Nuclear War, Bioweapons May Trigger End Of Humanity By 2100: Report
• AI agent integration lags behind adoption, creating security gaps
• used books — Mysterious Bulk Purchases in Korea Spark AI Training Fears