United States soldier Gannon Ken Van Dyke has been arrested and charged for placing bets on prediction marketplace Polymarket using classified information he had access to related to the capture of former Venezuelan president Nicolás Maduro . The US Army Special Forces master sergeant, who was directly involved with the planning and execution of the operation, allegedly made $409,881 in profits. According to the Department of Justice, Van Dyke created a Polymarket account around December 26, 2025 and made 13 bets related to Maduro from December 27 to January 2. He took the “Yes” position on several Polymarket wagers, including “US Forces in Venezuela… by January 31, 2026,” “Maduro out by… January 31, 2026, “Will the US invade Venezuela by January 31” and “Trump invokes War Powers against Venezuela by… January 31.” The US military captured Maduro and his wife on January 3. Van Dyke allegedly bet a total of $33,034 and made over ten times that amount from his winnings. He withdrew his money from Polymarket on the day Maduro was captured and then sent it to a foreign crypto vault before depositing it to a new online brokerage account. Shortly after Maduro’s capture, reports came out about how an anonymous gambler made almost half a million dollars before it was announced, raising concerns that someone had profited off insider military knowledge. The Justice Department says Van Dyke tried to cover his tracks. After reports about the potential insider bets were published, he allegedly asked Polymarket to delete his account, falsely claiming that he lost access to the email he used. He also changed the email address linked to his crypto account to another one not associated with his name. Van Dyke has been charged with three counts of violation against the Commodity Exchange Act, with each one carrying a max sentence of 10 years in prison. He has also been charged with one count of wire fraud with a max penalty of 20 years in prison, as well as one count of unlawful monetary transaction with a max sentence of 10 years. Prediction marketplaces have been struggling with insider trading problems, and this is far from the first incident. Recently, Kalshi took action against three political candidates, accusing them of insider trading related to their campaigns. Matt Klein of Minnesota and Ezekiel Enriquez of Texas face a fine of less than $1,000 and suspensions of up to five years. Meanwhile Mark Moran of Virginia faces disciplinary action, a five year suspension and a fine of more than $6,000. This article originally appeared on Engadget at https://www.engadget.com/apps/us-soldier-arrested-for-allegedly-making-over-400000-on-polymarket-with-classified-maduro-information-014531367.html?src=rss
DeepSeek has released its latest AI models, the V4 Pro and Flash versions, a bit over a year after it went viral and became the top rated free app on Apple's App Store in the US. “Welcome to the era of cost-effective 1 million context length,” DeepSeek said in its announcement . Context length is what you call the maximum number of tokens that an AI model can remember, so the bigger it is, the more coherent and consistent an AI is when it comes to extended conversations. OpenAI’s recently announced GPT‑5.5 has a context window ranging from 400,000 to 1 million, for instance. The new model is still open-source, allowing users to download its code and modify it if they want. DeepSeek says V4 Pro has enhanced agentic capabilities and claims that it rivals top closed-source models when it comes to reasoning. It also says that it trails only Gemini-3.1-Pro in rich world knowledge. Meanwhile, V4 Flash isn’t quite as powerful as the V4 Pro, but it has faster response times. Still, its reasoning abilities closely approach V4 Pro, DeepSeek says, and it performs on par with with the Pro version on simple Agent tasks. Shortly after DeepSeek topped the App Store charts, it was banned for use by US federal agencies and on government-owned devices. Authorities believed it was a national security risk and posed a threat to US AI stocks. South Korea also paused downloads of its app over privacy concerns. 🚀 DeepSeek-V4 Preview is officially live & open-sourced! Welcome to the era of cost-effective 1M context length. 🔹 DeepSeek-V4-Pro: 1.6T total / 49B active params. Performance rivaling the world's top closed-source models. 🔹 DeepSeek-V4-Flash: 284B total / 13B active params.… pic.twitter.com/n1AgwMIymu — DeepSeek (@deepseek_ai) April 24, 2026 This article originally appeared on Engadget at https://www.engadget.com/ai/deepseek-promises-its-new-ai-model-has-world-class-reasoning-115733512.html?src=rss
