ZEC今晨一度拉升突破1580美元，24小时涨超7%

Layer 1 blockchain Linera’s team released a statement saying that due to funding shortages, Linera will cease operations immediately and will not resume in the foreseeable future. The community round public offering launched earlier on the Sonar platform raised nearly $900,000 in commitments but failed to meet the minimum funding threshold, with all contributions fully refunded. Though the team later actively sought emergency funding to advance the project to mainnet, the effort ultimately fell short. Linera emphasized that the decision is unrelated to the technology itself, the team’s strength, or its community, and thanked all supporters. BlockBeats previously reported that on August 16, 2023, Linera completed a $6 million funding round led by Borderless Capital, with funds earmarked for expanding the team, launching the protocol’s devnet and testnet, building strategic influence in the Asia-Pacific region, and continuing to develop its developer academy. On June 29, 2022, Linera announced a $6 million seed round led by a16z, with participation from Cygni Capital, Kima Ventures, Tribe Capital, and others, which was intended to hire engineers and team members to build the protocol.

Beating AI Express News: Google confirmed that Gemini accidentally accessed the systems of three real companies during a cybersecurity evaluation. One gained entry via password guessing, while the other two found login credentials in public code repositories. Gemini stopped all actions immediately after identifying the targets as real companies. No sandbox escape took place throughout the process, as the public network entry was already open by default. The root cause lies with third-party evaluation firm Irregular: the test was not supposed to access the public network, but environment configuration errors allowed the model to connect directly to the internet. The simulated companies shared names with real ones, leading Gemini to follow the task trail and access external real systems. Irregular noted that this is the same issue that triggered similar incidents at other labs previously. Anthropic, OpenAI, and Meta have all encountered similar pitfalls in Irregular’s evaluations before. Anthropic and OpenAI both confirmed that the relevant incidents stemmed from test environments mistakenly enabling public network access; OpenAI further emphasized that there was no complex sandbox escape or zero-day vulnerability exploitation. Independent media outlet Effort criticized that these incidents were later framed as "model out-of-control" narratives.

According to EmberCN’s monitoring, a whale or institutional entity withdrew and hoarded 112,066 ETH from Bitfinex three years ago. At that time, ETH traded at around $2,030, resulting in a total holding cost of approximately $227 million. With ETH recently breaking above $2,600, the entity’s cumulative unrealized profit once hit roughly $66.45 million, a return of about 29%. About two hours ago, the entity transferred 21,229 ETH (valued at approximately $55.93 million) held in one of its addresses to Bitfinex, likely executing a partial profit-taking operation.

According to monitoring by EmberCN, the TRUMP token team address transferred out 11.25 million TRUMP tokens 12 days ago, valued at approximately $26 million. Of that amount, 3.25 million TRUMP tokens (worth around $6.9 million) were further transferred into OKX 7 hours ago. It remains unclear whether the relevant tokens will be sold.

The US Senate this week held a procedural vote that failed to advance the crypto market structure bill, the Clarity Act, with 49 votes in favor and 50 against. The bill remains far short of the 60-vote threshold required for passage. According to reports, all Democratic senators voted against the measure, joined by Republican senators Susan Collins, Josh Hawley, and Jerry Moran. Thom Tillis initially voted in favor before switching to no to preserve the option of reviving the bill in future. Multiple Democratic senators involved in negotiations described the vote as "a setback, not the end", noting that both parties have already begun efforts to restart talks. With congressional progress stalled, the crypto industry is turning more attention to regulatory bodies. SEC Chair Paul Atkins explicitly linked the agency’s innovation exemption measures announced this week to the Clarity Act’s setback, a move that paves the way for tokenized U.S. stocks to trade on-chain. Separately, CFTC staff issued a "no-action" position targeting passive software service providers, and submitted a broader crypto market rule proposal to the White House for review, with specific details remaining undisclosed.

Beating AI Express: Google said its Gemini AI model accidentally accessed the internet and gained access to the systems of three companies during a cybersecurity capability test, marking the first known case of a Google AI system autonomously performing such an operation. The incident occurred in May during a test conducted by cybersecurity firm Irregular. During the test, Gemini was performing tasks in a "capture the flag" (CTF) cybersecurity exercise, but it accessed real corporate systems because the virtual company names in the test environment matched real ones, plus the model was accidentally granted internet access. Google stated that in one instance, the model accessed a protected system by attempting passwords; in the other two cases, it found credentials in public online repositories and tried to access the corresponding systems. After confirming it had accessed real corporate systems, the model halted all further actions. Google noted the incident caused no harm to the affected companies, so it was not classified as an AI model out-of-control incident, adding that safety mechanisms helped the model stop its actions in time. The affected companies have been notified, and Google has also reported the incident to relevant regulatory authorities. The incident is similar to previous security test events disclosed by AI firms including OpenAI and Anthropic, and has once again drawn external attention to AI models’ ability to autonomously perform network operations and their safety protection mechanisms.