Google's Gemini AI model has been cited in a crypto investigator's report alleging that Tronify.rent took $69,651 from approximately 80 victims during September 2025. According to JP online of IOC Investigations, at least one victim reportedly asked Gemini whether the website was legitimate before interacting with it, and received a response that appeared to endorse the service. The investigator claims this represents a serious security concern, as independent threat researchers had already flagged Tronify.rent months earlier for phishing and brand impersonation risks. PhishDestroy recorded abuse reports against the domain in April while it remained publicly reachable, assigning it a critical threat rating of 90/100. The website currently claims to operate as a noncustodial TRON energy rental service, but victim reports tell a different story with funds being drained after connecting wallets to the platform.
As Alphabet Inc.'s Google prepares for the coming launch of Gemini 4, it's grappling with internal skepticism over how well the flagship artificial intelligence model performs in key areas, such as coding. According to Business Standard, while Gemini 4 has performed well on benchmarks the industry uses to gauge model efficacy, it does less well when employees actually put it to work. The model struggles to handle certain coding tasks, with employees reporting that Gemini isn't particularly adept at front-end design, which shapes how apps and websites look and feel. This represents a potentially serious setback as Google has struggled to compete in the red-hot market for AI coding tools. A Google employee familiar with model development said there is "large consensus" internally at the company that Gemini 4 is at the frontier, though some employees believe Anthropic's Fable and OpenAI's Astra models are improving at a faster rate.
Google has restricted access to its new Gemini 4 Argon AI model to a vetted group of cybersecurity experts due to safety concerns and potential misuse by hackers. As reported by The Straits Times, Koray Kavukcuoglu, Google's chief AI architect, announced in a blog post that "Safely releasing frontier capabilities at this level requires a phased approach." The company is voluntarily giving the US government early access to the model and will gather feedback from testers before making it widely available. This cautious approach mirrors the strategy of rival Anthropic, which has kept its most advanced model, Claude Mythos Preview, restricted to a few trusted organizations. The announcement came after President Donald Trump hosted top tech executives, including Google chief Sundar Pichai and Anthropic's Dario Amodei, at the White House, where they signed a voluntary accord pledging to police AI system risks.
Google's latest benchmark data reveals that 'Argon' leads or ties for first place in 13 of 18 disclosed categories, establishing it as the leading frontier model by total number of benchmark leads. According to VentureBeat, the model posts the highest score or ties for first on 12 benchmarks, while GPT-6 Astra leads outright on 3 benchmarks and Claude Opus 5.5 leads on 2. In critical enterprise workflows, Argon scores 19.6% on Harvey's Legal Agent Benchmark, significantly ahead of GPT-6 Astra at 5.4% and Claude Opus 5.5 at 3.8%. On AutomationBench for business execution, Argon achieves 51.3% compared to Claude Opus 5.5's 42.5% and GPT-6 Astra's 41.4%. The model demonstrates strong performance in long-video understanding with 91.7% on LVBench and financial analysis with 65.4% on Vals Finance Agent v2. As per The Straits Times, Google said Argon excels at complex tasks in software engineering, legal and financial work and cyber defence, with a leading ability to find and fix critical software flaws. Early testers used Argon to uncover a flaw in software used by hospitals around the world that exposed sensitive personal information - something other advanced models had missed.
Google has laboriously developed models that can compete with OpenAI and Anthropic PBC, with the company having planned to release a different version in June, dubbed Gemini 3.5 Pro, but abandoned the effort. According to Business Standard, the decision likely cost the company dearly in time and money, with training runs to build such a model requiring spending as much as $400 million, according to Bloomberg Intelligence analyst Mandeep Singh. The company is encountering various challenges with Gemini 4, with its coding abilities being uneven and the model being very large, according to a person familiar with its development. Typically big models are expensive to run, potentially putting pressure on Google's margins. Experts say Google could be suffering from an industry tendency to focus on benchmarks - a phenomenon known as "benchmaxxing," when engineers concentrate more on achieving a good score than creating a product that does a job well. Google explicitly warns that generative AI can produce inaccurate information and tells users to check claims presented as facts using other sources before acting on AI-generated responses. The company's AI security documentation explains that malicious webpages can contain material intended to influence generative AI systems through prompt-injection techniques, with Gemini using safeguards to detect suspicious material.
Google badly needs Gemini 4 to succeed as versions of the model underpin nearly every product the company sells, from the AI answers atop Search, Google's main profit engine, to Maps, Gmail and Chrome. Each of those products has more than a billion users, a distribution advantage some of the company's rivals lack. However, OpenAI and Anthropic are increasingly moving beyond selling models to building products of their own, including coding agents. Failing to deliver a cutting-edge model could give Google's competitors more time to convince consumers, developers and businesses that the future of search and software should run on their platforms instead. In response to questions from Bloomberg, Google said that even though its last Pro model was released in February, the company has since seen growth in its AI products, including the enterprise version of Gemini as well as its consumer chatbot app and AI Mode in Google Search, the last two of which have crossed 1 billion users. The Tronify allegations highlight the growing risk of AI being used as part of crypto scam tactics, with security researchers documenting increasingly sophisticated campaigns that use trusted technology platforms to make malicious offers appear safer.
