Author: admin

AI models are being cranked out at a dizzying pace, by everyone from Big Tech companies like Google to startups like OpenAI and Anthropic. Keeping track of the latest ones can be overwhelming.  Adding to the confusion is that AI models are often promoted based on industry benchmarks. But these technical metrics often reveal little about how real people and companies actually use them.  To cut through the noise, TechCrunch has compiled an overview of the most advanced AI models released since 2024, with details on how to use them and what they’re best for. We’ll keep this list updated…

The US Justice Department on Wednesday unsealed indictments charging employees of a Chinese cybersecurity firm known as i-Soon (Anxun Information Technology) with conducting extensive hacking campaigns on behalf of Beijing’s security services. Prosecutors say i-Soon employees acted as “hackers-for-hire,” breaching email networks, government databases, and corporate systems at the direction of China’s Ministry of Public Security (MPS) and the Ministry of State Security (MSS). The indictments come a year after an unauthorized and highly unusual online dump of documents from i-Soon that catalog apparent hacking activity and tools to spy on both Chinese and foreigners. Among the apparent targets of…

Gong, a startup that helps companies predict their revenue from potential sales, has surpassed $300 million in annualized recurring revenue, the company announced on Wednesday. Since its founding in 2016, Gong has used AI to analyze customer interactions. The addition of generative AI capabilities in recent years has helped to fuel the company’s growth. “We’re seeing great momentum. That’s why we’re excited to share the numbers,” Gong CEO Amit Bendov told TechCrunch. Gong was last valued at $7.25 billion when it raised a $250 million Series E in 2021 in a deal led by Franklin Templeton with participation from Coatue,…

Prime Video is now experimenting with AI-assisted dubbing for select licensed movies and TV shows, as announced by the Amazon-owned streaming service on Wednesday. According to Prime Video, this new test will feature AI-assisted dubbing services in English and Latin American Spanish, combining AI with human localization professionals to “ensure quality control,” the company explained.   Initially, it’ll be available for 12 titles that previously lacked dubbing support. These titles include the Spanish animated film “El Cid: La Leyenda,” the family-friendly drama “Mi Mamá Lora,” and the indie movie “Long Lost.” AI-based dubbing has gained popularity in the entertainment industry as…

NEW YORK (AP) — Barnard College’s library was evacuated Wednesday as police responded to a bomb threat during a sit-in staged by pro-Palestinian protesters.The New York Police Department said on X, formerly Twitter, that a bomb threat was reported at the upper Manhattan college’s Milstein Center, which serves as the hub for academic life on campus.The department said anyone refusing to leave the building during the evacuation would be subject to arrest. “Please stay away from the area,” the police said in its post.The department said later that officers were still on the scene and that arrests had been made.…

As part of the iOS 18.4 software update, currently in public beta, Apple is introducing AI-powered summaries of App Store reviews. The new feature will leverage Apple Intelligence, the company’s built-in AI technology, to offer an overall summary based on the reviews others have left on the App Store. The review summaries will be generated by large language models (LLMs) and will highlight key information into a short paragraph, Apple’s website explains. The summaries will also be refreshed weekly for apps and games that have enough reviews to generate a summary — though Apple did not say what that threshold…

OpenAI has begun rolling out its newest AI model, GPT-4.5, to users on its ChatGPT Plus tier. In a series of posts on X, OpenAI said that the rollout will take “1-3 days,” and that it expects rate limits to change. GPT-4.5 launched first for subscribers to OpenAI’s $200-a-month ChatGPT Pro plan last week. “We’d like to give everyone with access to GPT-4.5 a sizable rate limit, but we expect rate limits to change as we learn more about demand,” the company wrote in its post. GPT-4.5 is OpenAI’s largest AI model yet, trained using more computing power and data…

Kin, a direct-to-consumer digital home insurance provider, has officially launched its home insurance offering across the entire state of California.California’s homeowners have struggled for years to secure comprehensive insurance, particularly in regions prone to wildfires, Kin explained. Even prior to the devastating January 2025 wildfires in Los Angeles County, the state had seen insurers scaling back their exposure.Reports from the California Senate indicated that nonrenewal rates had tripled in the previous year, leaving tens of thousands of residents without coverage. Many policyholders were either forced to go without insurance altogether or resort to the California Fair Access to Insurance Requirements…

DXC Technology, a Fortune 500 global technology services provider, and ServiceNow, a leading AI-powered platform for business transformation, have expanded their strategic partnership to modernise operations across the insurance industry.The partnership was formed in response to growing pressures on insurers to accelerate innovation and growth while improving operational agility. Many insurance firms are still reliant on manual processes that slow down workflows and create inefficiencies. According to research from HFS Research, 45% of insurers are already investing in technology aimed at aligning front, middle, and back-office operations.Together, DXC and ServiceNow have developed DXC Assure BPM powered by ServiceNow, a comprehensive…

Many organizations are still concerned that patching operational technology (OT) systems can lead to equipment downtime and operational disruptions, and consequently they do not conduct regular patching, according to cyber-physical security firm TXOne Networks.  The data comes from TXOne’s 2024 Annual OT/ICS Cybersecurity Report, which is based on a survey of 150 C-level executives in North America, Europe, the Middle East and Asia.  The survey found that 85% of organizations don’t conduct regular patching. A majority install patches quarterly or less often, which leaves them exposed to attacks for extended periods of time.  This is despite a vast majority experiencing…