By PK Chakrabarti ( January 15, 2025, 2:57 PM EST) — The new federal government will likely bring significant changes in U.S. patent policy. These changes, spanning leadership transitions at the U.S. Patent and Trademark Office, industry policies, legislative initiatives and international trade strategies, will reflect the government’s renewed focus on strengthening intellectual property rights, fostering innovation and enhancing the nation’s competitive edge…. Law360 is on it, so you are, too. A Law360 subscription puts you at the center of fast-moving legal issues, trends and developments so you can act with speed and confidence. Over 200 articles are published daily across more than 60 topics, industries, practice areas and jurisdictions. A Law360 subscription includes features such as Daily newsletters Expert analysis Mobile app Advanced search Judge information Real-time alerts 450K+ searchable archived articles And more! Experience Law360 today with a free 7-day trial. source

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More Researchers at the Mohamed bin Zayed University of Artificial Intelligence (MBZUAI) have announced the release of LlamaV-o1, a state-of-the-art artificial intelligence model capable of tackling some of the most complex reasoning tasks across text and images. By combining cutting-edge curriculum learning with advanced optimization techniques like Beam Search, LlamaV-o1 sets a new benchmark for step-by-step reasoning in multimodal AI systems. “Reasoning is a fundamental capability for solving complex multi-step problems, particularly in visual contexts where sequential step-wise understanding is essential,” the researchers wrote in their technical report, published today. Fine-tuned for reasoning tasks that require precision and transparency, the AI model outperforms many of its peers on tasks ranging from interpreting financial charts to diagnosing medical images. In tandem with the model, the team also introduced VRC-Bench, a benchmark designed to evaluate AI models on their ability to reason through problems in a step-by-step manner. With over 1,000 diverse samples and more than 4,000 reasoning steps, VRC-Bench is already being hailed as a game-changer in multimodal AI research. LlamaV-o1 outperforms competitors like Claude 3.5 Sonnet and Gemini 1.5 Flash in identifying patterns and reasoning through complex visual tasks, as demonstrated in this example from the VRC-Bench benchmark. The model provides step-by-step explanations, arriving at the correct answer, while other models fail to match the established pattern. (credit: arxiv.org) How LlamaV-o1 stands out from the competition Traditional AI models often focus on delivering a final answer, offering little insight into how they arrived at their conclusions. LlamaV-o1, however, emphasizes step-by-step reasoning — a capability that mimics human problem-solving. This approach allows users to see the logical steps the model takes, making it particularly valuable for applications where interpretability is essential. The researchers trained LlamaV-o1 using LLaVA-CoT-100k, a dataset optimized for reasoning tasks, and evaluated its performance using VRC-Bench. The results are impressive: LlamaV-o1 achieved a reasoning step score of 68.93, outperforming well-known open-source models like LlaVA-CoT (66.21) and even some closed-source models like Claude 3.5 Sonnet. “By leveraging the efficiency of Beam Search alongside the progressive structure of curriculum learning, the proposed model incrementally acquires skills, starting with simpler tasks such as [a] summary of the approach and question derived captioning and advancing to more complex multi-step reasoning scenarios, ensuring both optimized inference and robust reasoning capabilities,” the researchers explained. The model’s methodical approach also makes it faster than its competitors. “LlamaV-o1 delivers an absolute gain of 3.8% in terms of average score across six benchmarks while being 5X faster during inference scaling,” the team noted in its report. Efficiency like this is a key selling point for enterprises looking to deploy AI solutions at scale. AI for business: Why step-by-step reasoning matters LlamaV-o1’s emphasis on interpretability addresses a critical need in industries like finance, medicine and education. For businesses, the ability to trace the steps behind an AI’s decision can build trust and ensure compliance with regulations. Take medical imaging as an example. A radiologist using AI to analyze scans doesn’t just need the diagnosis — they need to know how the AI reached that conclusion. This is where LlamaV-o1 shines, providing transparent, step-by-step reasoning that professionals can review and validate. The model also excels in fields like chart and diagram understanding, which are vital for financial analysis and decision-making. In tests on VRC-Bench, LlamaV-o1 consistently outperformed competitors in tasks requiring interpretation of complex visual data. But the model isn’t just for high-stakes applications. Its versatility makes it suitable for a wide range of tasks, from content generation to conversational agents. The researchers specifically tuned LlamaV-o1 to excel in real-world scenarios, leveraging Beam Search to optimize reasoning paths and improve computational efficiency. Beam Search allows the model to generate multiple reasoning paths in parallel and select the most logical one. This approach not only boosts accuracy but reduces the computational cost of running the model, making it an attractive option for businesses of all sizes. LlamaV-o1 excels in diverse reasoning tasks, including visual reasoning, scientific analysis and medical imaging, as shown in this example from the VRC-Bench benchmark. Its step-by-step explanations provide interpretable and accurate outcomes, outperforming competitors in tasks such as chart comprehension, cultural context analysis and complex visual perception. (credit: arxiv.org) What VRC-Bench means for the future of AI The release of VRC-Bench is as significant as the model itself. Unlike traditional benchmarks that focus solely on final answer accuracy, VRC-Bench evaluates the quality of individual reasoning steps, offering a more nuanced assessment of an AI model’s capabilities. “Most benchmarks focus primarily on end-task accuracy, neglecting the quality of intermediate reasoning steps,” the researchers explained. “[VRC-Bench] presents a diverse set of challenges with eight different categories ranging from complex visual perception to scientific reasoning with over [4,000] reasoning steps in total, enabling robust evaluation of LLMs’ abilities to perform accurate and interpretable visual reasoning across multiple steps.” This focus on step-by-step reasoning is particularly critical in fields like scientific research and education, where the process behind a solution can be as important as the solution itself. By emphasizing logical coherence, VRC-Bench encourages the development of models that can handle the complexity and ambiguity of real-world tasks. LlamaV-o1’s performance on VRC-Bench speaks volumes about its potential. On average, the model scored 67.33% across benchmarks like MathVista and AI2D, outperforming other open-source models like Llava-CoT (63.50%). These results position LlamaV-o1 as a leader in the open-source AI space, narrowing the gap with proprietary models like GPT-4o, which scored 71.8%. AI’s next frontier: Interpretable multimodal reasoning While LlamaV-o1 represents a major breakthrough, it’s not without limitations. Like all AI models, it is constrained by the quality of its training data and may struggle with highly technical or adversarial prompts. The researchers also caution against using the model in high-stakes decision-making scenarios, such as healthcare or financial predictions, where errors could have serious consequences. Despite these challenges, LlamaV-o1 highlights the growing importance of multimodal AI systems that can seamlessly integrate text, images

By Catherine Marfin ( January 14, 2025, 8:26 PM EST) — Texas’ push before the U.S. Supreme Court for a relaxed standard of judicial review in First Amendment cases is unlikely to come to fruition, as decades of precedent work against the state’s law requiring age verification on pornography sites…. Law360 is on it, so you are, too. A Law360 subscription puts you at the center of fast-moving legal issues, trends and developments so you can act with speed and confidence. Over 200 articles are published daily across more than 60 topics, industries, practice areas and jurisdictions. A Law360 subscription includes features such as Daily newsletters Expert analysis Mobile app Advanced search Judge information Real-time alerts 450K+ searchable archived articles And more! Experience Law360 today with a free 7-day trial. source

The majority of organizations — 89% of them, according to the 2024 State of the Cloud Report from Flexera — have adopted a multicloud strategy. Now they are riding the wave of the next big technology: AI. The opportunities seem boundless: chatbots, AI-assisted development, cognitive cloud computing, and the list goes on. But the power of AI in the cloud is not without risk.   While enterprises are eager to put AI to use, many of them still grapple with data governance as they accumulate more and more information. AI has the potential to amplify existing enterprise risks and introduce entirely new ones. How can enterprise leaders define these risks, both internal and external, and safeguard their organizations while capturing the benefits of cloud and AI?   Defining the Risks   Data is the lifeblood of cloud computing and AI. And where there is data, there is security risk and privacy risk. Misconfigurations, insider threats, external threat actors, compliance requirements, and third parties are among the pressing concerns enterprise leaders must address  Risk assessment is not a new concept for enterprise leadership teams. Many of the same strategies apply when evaluating the risks associated with AI. “You do threat modeling and your planning phase and risk assessment. You do security requirement definitions [and] policy enforcement,” says Rick Clark, global head of cloud advisory at UST, a digital transformations solutions company.   Related:Are We Ready for Artificial General Intelligence? As AI tools flood the market and various business functions clamor to adopt them, the risk of exposing sensitive data and the attack surface expands.   For many enterprises, it makes sense to consolidate data to take advantage of internal AI, but that is not without risk. “Whether it’s for security or development or anything, [you’re] going to have to start consolidating data, and once you start consolidating data you create a single attack point,” Clark points out.   And those are just the risks security leaders can more easily identify. The abundance of cheap and even free GenAI tools available to employees adds another layer of complexity.   “It’s [like] how we used to have the shadow IT. It’s repeating again with this,” says Amrit Jassal, CTO at Egnyte, an enterprise content management company.   AI comes with novel risks as well.   “Poisoning of the LLMs, that I think is one of my biggest concerns right now,” Clark shares with InformationWeek. “Enterprises aren’t watching them carefully as they’re starting to build these language models.”  Related:AI’s on Duty, But I’m the One Staying Late How can enterprises ensure the data feeding the LLMs they use hasn’t been manipulated?  This early on in the AI game, enterprise teams are faced with the challenges of a managing the behavior and testing systems and tools that they may not yet fully understand.   “What’s … new and difficult and challenging in some ways for our industry is that the systems have a kind of nondeterministic behavior,” Mark Ryland, director of Amazon Security for Amazon Web Services (AWS), explains. “You can’t comprehensively test a system because it’s designed in part to be critical, creative, meaning that the very same input doesn’t result in the same output.”  The risks of AI and cloud can multiply with the complexity of an enterprise’s tech stack. With a multi-cloud strategy and often growing supply chain, security teams have to think about a sprawling attack surface and myriad points of risk.   “As an example, we have had to take a close look at least privilege things, not just for our customers but for our own employees as well. And, then that has to be extended not to just one provider but to multiple providers,” says Jassal. “It definitely becomes much more complex.”  AI Against the Cloud  Widely available AI tools will be leveraged not only by enterprises but also the attackers that target them. At this point, the threat of AI-fueled attacks on cloud environments is moderately low, according to IBM’s X-Force Cloud Threat Landscape Report 2024. But the escalation of that threat is easy to imagine.   Related:How Do Companies Know if They Overspend on AI and Then Recover? AI could exponentially increase threat actors’ capabilities via coding-assistance, increasingly sophisticated campaigns, and automated attacks.   “We’re going to start seeing that AI can gather information to start making … personalized phishing attacks,” says Clark. “There’s going to be adversarial AI attacks, where they exploit weaknesses in your AI models even by feeding data to bypass security systems.”  AI model developers will, naturally, attempt to curtail this activity, but potential victims cannot assume this risk goes away. “The providers of GenAI systems obviously have capabilities in place to try to detect abusive use of their systems, and I’m sure those controls are reasonably effective but not perfect,” says Ryland.   Even if enterprises opt to eschew AI for now, threat actors are going to use that technology against them. “AI is going to be used in attacks against you. You’re going to need AI to combat it, but you need to secure your AI. It’s a bit of a vicious circle,” says Clark.   The Role of Cloud Providers  Enterprises still have responsibility for their data in the cloud, while cloud providers play their part by securing the infrastructure of the cloud.   “The shared responsibility still stays,” says Jassal. “Ultimately if something happens, a breach etcetera, in Egnyte’s systems … Egnyte is responsible for it whether it was due to a Google problem or Amazon problem. The customer doesn’t really care.”  While that fundamental shared responsibility model remains, does AI change that conversation at all?  Model providers are now part of the equation. “Model providers have a distinct set of responsibilities,” says Ryland. “Those entities … [take] on some responsibility to ensure that the models are behaving according to the commitments that are made around responsible AI.”  While different parties — users, cloud providers, and model providers — have different responsibilities, AI is giving them new ways to meet those responsibilities.   AI-driven security, for example, is going to be essential for enterprises to protect their

By Grace Elletson ( January 15, 2025, 4:53 PM EST) — A federal judge declined to cut a supervisor from a former software company worker’s suit claiming he was fired after his boss found out he is gay, stating the Arizona-based supervisor can still be sued in Michigan even though he managed the ex-employee remotely…. Law360 is on it, so you are, too. A Law360 subscription puts you at the center of fast-moving legal issues, trends and developments so you can act with speed and confidence. Over 200 articles are published daily across more than 60 topics, industries, practice areas and jurisdictions. A Law360 subscription includes features such as Daily newsletters Expert analysis Mobile app Advanced search Judge information Real-time alerts 450K+ searchable archived articles And more! Experience Law360 today with a free 7-day trial. source

The Linux Foundation announced on Thursday a fund to support and organize open projects to build Chromium-based browsers. The Supporters of Chromium-Based Browsers now accepts new members. “With the launch of the Supporters of Chromium-Based Browsers, we are taking another step forward in empowering the open source community,” said Jim Zemlin, executive director of the Linux Foundation, in a press release. “This project will provide much-needed funding and development support for open development of projects within the Chromium ecosystem.” Many browsers, including Arc, Microsoft Edge, and Opera, use Google’s Chromium as their underlying infrastructure. Interested potential partners can apply with the Linux Foundation. What’s hot at TechRepublic What is the Supporters of Chromium-Based Browsers group? The purpose of the Supporters of Chromium-Based Browsers group is to “provide a neutral space where industry leaders, academia, developers, and the broader open source community can work together to support projects within the Chromium ecosystem,” the Linux Foundation wrote. As such, the group will “remove barriers to innovation, expand adoption, and ensure that projects within the Chromium ecosystem receive the resources they need to thrive.” Google, Meta, Microsoft, and Opera have endorsed the Supporters of Chromium-Based Browsers. The initiative will be a “neutral space” for “industry leaders, academia, developers, and the broader open source community,” the Linux Foundation said. Nothing will change about existing Chromium products as a result of this announcement, according to the Linux Foundation. Instead, the Supporters of Chromium-Based Browsers will be a new group within the Linux Foundation, following an “open governance model,” the foundation said. In addition, a technical advisory committee will shape the development of the initiative and align it with the needs of the larger Chromium community. “This initiative aligns with our commitment to the web platform through meaningful and positive contributions, engagement in collaborative engineering, and partnerships with the community to achieve the best outcome for everyone using the web,” said Meghan Perez, vice president of Microsoft Edge, in the press release. SEE: The U.K.’s competition regulator probed Apple in November regarding the dominance of the WebKit browser engine required on iOS. “With the incredible support of the Linux Foundation, we believe the Supporters of Chromium-Based Browsers is an important opportunity to create a sustainable platform to support industry leaders, academia, developers, and the broader open source community in the continued development and innovation of the Chromium ecosystem,” said Parisa Tabriz, vice president of Chrome, in the press release. Google will continue to contribute to Chromium Chromium was released alongside proprietary browser Chrome in 2008 as an open-source framework option. The creation of the Supporters of Chromium-Based Browsers comes within months of the U.S. government ordering Google to divest from Chrome to prevent a monopoly. While the Supporters initiative is not linked directly to the divestiture, it does provide some structure for Chromium going forward. Google makes the majority (about 94 percent) of contributions to Chromium, according to a Jan. 9 blog from Chrome. “Others” and the open source cooperative Igalia are distant second and third top contributors. “While we have no intention of reducing this investment, we continue to welcome others stepping up to invest more,” the blog states. Those contributions include running thousands of servers and responding to hundreds of bugs per day, Chrome said. source

The mobility industry is pioneering the use of data spaces   The mobility industry is at the forefront of the data spaces initiative. Almost 180 players are currently involved in its data space Catena-X – major manufacturers as well as suppliers, mobility providers and organizations such as ADAC. Their stated goal: a common data infrastructure for the entire supply chains of the industry.   Raw material suppliers, component suppliers, carmakers, users and recyclers are also networking via Catena-X to better trace sources of error. So far, companies have only documented information on processes and components for which they are responsible internally. In the future, end-to-end data chains will show exactly who has installed which materials and components or which software has been used.   A second shared data space for the industry is the aforementioned MDS, which is funded by the Austrian Federal Ministry for Digital and Transport (BMDV). Around 150 companies are currently active there – mobility service providers, municipalities, public transport companies, companies from the energy sector, mobile network operators and research institutions. The data offered by MDS includes information on traffic and traffic flow, construction sites and road conditions, parking and public transport, car and bike sharing and much more.   source

By Lauren Berg ( January 14, 2025, 7:27 PM EST) — Elon Musk violated securities laws by failing to timely disclose his initial buy-up of Twitter stock ahead of his $44 billion acquisition of the company, allowing him to purchase shares at artificially low prices, the U.S. Securities and Exchange Commission alleged in a D.C. federal lawsuit filed Tuesday…. Law360 is on it, so you are, too. A Law360 subscription puts you at the center of fast-moving legal issues, trends and developments so you can act with speed and confidence. Over 200 articles are published daily across more than 60 topics, industries, practice areas and jurisdictions. A Law360 subscription includes features such as Daily newsletters Expert analysis Mobile app Advanced search Judge information Real-time alerts 450K+ searchable archived articles And more! Experience Law360 today with a free 7-day trial. source

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More ChatGPT is taking a significant step toward becoming a full-blown personal assistant with the release of a new feature called Tasks. This could signal that OpenAI will release more agents in the future. Currently in beta, Tasks lets ChatGPT Plus, Team and Pro users schedule actions ahead of time. For example, if someone wants to receive project reminders or daily weather updates, they can prompt ChatGPT, which will notify them of the chosen date and time. Tasks can be recurring or one-time reminders.  To set up a task, users should toggle to “4o with scheduled tasks” on the model picker and write a reminder prompt. ChatGPT can also suggest tasks from previous conversations. Tasks work with all versions of ChatGPT and send notifications through desktop, web and mobile. However, users can access the task manager only on the web version of ChatGPT.  OpenAI said the beta period will help its researchers “understand how people use Tasks and refine the feature before making it available to all ChatGPT users.” Tasks join other assistant-like features for ChatGPT. During its “12 Days of OpenAI” event in December, OpenAI launched screen sharing, letting users open ChatGPT while reading a text message and asking it to help them respond. OpenAI’s first agent? Rumors around OpenAI releasing an AI agent swirled when some users caught ChatGPT providing access to scheduled tasks as far back as December. The agent, called Operator, would be the company’s first agent.  People believed Tasks could be the precursor to Operator. X user @kimmonismus, aka “Chubby,” said, “The Information seems to be right and everything is being prepared for the release of ‘Operator,’ i.e. OpenAI’s agent. The ‘Tasks’ function found by Tibor seems to be the first significant step in the preparation of Operator. It is questionable whether we will get a release this month or just a preview.” Testing Catalog News on X theorized Tasks could eventually enable ChatGPT to search for specific information, summarize data, open websites, access documents and think through problems. VentureBeat reached out to OpenAI about Tasks and Operator. The company declined to answer the question and only said Tasks “will be an important step toward making ChatGPT a more helpful AI companion.” OpenAI has already made its first foray into the agentic space with Swarm, a framework it released to help orchestrate AI agents.  Making ChatGPT a better assistant  If you’re like me, you probably use the Reminders app and Google Calendar too much to remind yourself to text a relative Happy Birthday or of your brunch plans, or to alert you to news embargoes and planned interviews.  There are many reminder, calendar and productivity apps available for both consumers and enterprises. These include Google Calendar, Outlook Calendar, Asana, Trello and Notion. The productivity assistant space is not short of applications that remind individuals and teams of tasks they must accomplish. OpenAI’s​​ big play in such a crowded space is interesting, especially since most people don’t think of chatbots as scheduling assistants. But ChatGPT already streamlines the process for users to migrate their coding or writing tasks to the platform or search the web without leaving the chat interface. ChatGPT even opens up developers’ IDE nearly automatically for them.  As ChatGPT adds more actions to its platform, setting up scheduled tasks and reminders is not so far-fetched. This makes ChatGPT a viable competitor to many productivity and scheduling apps.  source

The wildfires in Los Angeles have torn through roughly 30,000 acres so far, devastating multiple communities. This inferno is the most destructive natural disaster in LA County’s history. With median home values surpassing $2 million, estimates indicate that insured losses may reach $20 billion. The widespread destruction is further worsening the state’s existing insurance crisis of homeowners facing unaffordable premiums or no coverage. Three factors have led to this issue: Premiums were too low. Prior to facing the brunt of wildfires, home insurance rates in California were relatively low. In Pacific Palisades, the area hit hardest by wildfires, premiums have been cheaper than in 97% of zip codes in the US. The frequency and severity of wildfires have increased. As a result, losses from them have become so costly that insurance companies — a homeowner’s last line of defense when disaster strikes — have canceled policies, raised rates, or exited the state. State Farm, the largest insurer in the state, has dropped 72,000 policies since March of 2024. California’s regulatory environment was too consumer-friendly. The regulatory environment has kept insurance premiums relatively low, even in high-risk areas. While the insurance regulations protect consumers from sudden rate hikes, this also limits the ability of insurers to respond to increasing risks and costs. Only recently has the state’s insurance regulator permitted double-digit rate increases, but this happened only after years of friction between regulators and the insurance industry. Lack of adequate insurance resulted in fewer, often cost-prohibitive coverage options. Homeowners have largely replaced their fire coverage with a last-resort state plan. The California FAIR Plan’s potential exposure to losses surged by 61% last year. In Pacific Palisades, the number of residential policies under the FAIR Plan grew by 85% to 1,430 last year. State regulations stipulate that the FAIR Plan can turn to the private insurance companies operating in the state to fill any gap to cover fire claims. Home insurers in the Golden State will need to adjust their operating models in the following ways: Understand and appropriately price for the spiraling costs of natural disasters. The impact of intense climate change has left insurers attempting to respond to risks that they do not fully understand. Setting premiums excessively high to create a margin for error has made insurance unaffordable. Insurers must invest in understanding the hazards and vulnerabilities associated with wildfire exposure to increase their underwriting accuracy. Utilize predictive modeling techniques. This will allow insurers to segment and price risks according to anticipated future damages. In California and in most other states, regulators have recently allowed insurers to set rates based on these models. Rather than only use actuarial models that are based on past loss history, insurers must leverage data science and modern analytical techniques to price risks more accurately. Hedge risks through reinsurance. Insurance carriers can protect themselves from the higher layers of risk exposure through risk transfer to the reinsurance ecosystem. While reinsurance rates for risks tied to climate change have increased in recent years, the changing regulatory environment allows insurers to transfer some of this over to the reinsurer. Insurers must leverage the reinsurance broker community to help build the appropriate reinsurance structures for disciplined portfolio management. Strike a balance with regulators. The California Department of Insurance must balance the need to protect consumers from evolving risk at affordable prices while ensuring that insurance companies are attracted to their state and do not become insolvent. The industry and the regulators must work hand in hand to create appropriate solutions at fair prices for consumers and insurers. The insurance industry exists to provide resiliency to societies, especially during extreme events and natural disasters. Without it, communities will destabilize in the face of increasing natural disasters. A solution for California exists. It will take the cooperation of regulators, insurers, and consumers to develop it. Clients interested in discussing this topic can chat with me via an inquiry or guidance session. source