Latest news with #AIgovernance
Associated Press
30-07-2025
- Business
- Associated Press
Geneva to host global foundation to oversee ethical AI
Initiatives to regulate AI are multiplying, such as the future GAIGI foundation, led by Geneva-based Axel Mazolo. Led by Geneva-based AI governance expert Axel Mazolo, the Geneva AI Governance Institute (GAIGI), scheduled to be launched in fall 2025, is an initiative that could mark a strategic milestone in the global regulation of so-called sensitive technologies. According to the initiator, sensitive technology or AI refers to an artificial intelligence system whose decision-making or potential malfunctions could have a significant impact on health, safety, individual freedoms, or the continuity of vital services. In a context where AI applications are multiplying in defense, healthcare, and critical infrastructure, this foundation aims to respond to a growing need for standards. Ensuring the reliability and security of artificial intelligence The first of its kind, the foundation will not issue certifications directly. Instead, it will focus on accrediting certification bodies around the world based on technical, ethical, and governance standards developed in Geneva. This innovative approach aims to establish a neutral and transparent layer of regulation while ensuring ongoing quality control through regular audits. The challenge is to guarantee the reliability and security of so-called sensitive AI systems. In terms of its organization, GAIGI's governance is designed to reflect Swiss values of neutrality: a high-level board of directors bringing together influential figures from international NGOs, industry, defense, law, and academic research. In addition, there are specialized committees covering technical, ethical, and compliance aspects. All of this is under the exclusive supervision of the Geneva cantonal authority, with no possibility of delegation to federal or foreign authorities. Known for his stance on the dangers of military automation and his advocacy for ethical AI governance, Axel Mazolo has established himself as an influential voice in debates on sensitive applications of artificial intelligence. He has published and collaborated with various think tanks on algorithmic security policies. He promotes a vision in which Switzerland, and more specifically Geneva, can play a pivotal role by offering an independent frame of reference in the face of international regulatory fragmentation. Geneva: capital of sensitive AI regulation This announcement comes at a time when Geneva, which has just hosted several major events dedicated to artificial intelligence, is clearly demonstrating its ambition to play a central role in digital governance. The initiative is part of a broader strategy promoted by local authorities to make the city on Lake Geneva a global center for technological regulation. Swiss neutrality, combined with the local UN and academic ecosystem, offers a unique environment for structuring cross-cutting standards capable of transcending geopolitical rivalries. While the European Union, the United States, and China are adopting national frameworks that are sometimes contradictory, GAIGI proposes a middle ground: a trusted, neutral, flexible, and rigorous infrastructure. If this vision prevails, Geneva could become the global capital for regulating high-risk artificial intelligence—led by a local foundation rooted in the spirit of Geneva and determined to make technological governance a pillar of global security. For inquiries or to schedule an interview with Axel Mazolo, contact Maeva Paris at [email protected] or call +41 78 313 93 53. To learn more, visit Media Contact Company Name: Mazolo & Cie Contact Person: Maeva Paris Email: Send Email City: Geneva Country: Switzerland Website: Press Release Distributed by To view the original version on ABNewswire visit: Geneva to host global foundation to oversee ethical AI
Yahoo
27-07-2025
- Business
- Yahoo
DEEP Robotics Showcases Core Technologies & Real-World Applications of Embodied Intelligence Robots at WAIC 2025
SHANGHAI, CN / / July 27, 2025 / On July 26, the 2025 World Artificial Intelligence Conference (WAIC) and High-Level Meeting on Global AI Governance kicked off in grand fashion at the Shanghai World Expo Center. DEEP Robotics, a leading innovator in embodied intelligence technology and industrial applications, took center stage at the WAIC Expo Hall. For the first time, DEEP Robotics brought real-world operation scenarios of embodied intelligence into the exhibition, showcasing the full-process inspection capabilities of industrial-grade quadruped robots on site. The latest iterations of the Lite series robots, along with multiple industry application solutions, were unveiled simultaneously, demonstrating the company's leadership in both core technology breakthroughs and real-world implementations. This year's conference spans three days, with exhibition space exceeding 70,000 square meters for the first time, attracting over 800 leading enterprises both from home and abroad to Shanghai. Under the theme - Intelligent Connectivity for a Shared Future - the event presents a comprehensive ecosystem from AI core technologies to practical industry applications. Within its 150-square-meter immersive exhibition area, DEEP Robotics became a highlight of the event by recreating a real electric-power inspection scenario, performing live demonstrations of full-process operations, and showcasing the embodied intelligence capabilities of its newly iterated core products. Power Inspection Scenario Faithfully Recreated, Demonstrating Industrial-Grade Capabilities As a pioneer in industry-focused embodied intelligence solutions, DEEP Robotics built a dedicated industry demonstration zone at the exhibition, authentically recreating a power inspection scenario to vividly showcase the real-world operational capabilities of quadruped robots in industrial environments. In the simulated power inspection setup at the booth, X30 quadruped robot took on the role of an efficient inspector, accurately identifying inspection targets such as analog gauges, digital and status meters, infrared temperature readings, and indicator lights. To simulate the endurance demands of real industrial environments, the exhibition area was also equipped with a charging pile. X30 can autonomously monitor its battery level and navigate to the charging station when needed, enabling unattended continuous operation. This setup offered visitors a clear view of the feasibility of quadruped robots replacing manual labor in high-risk, high-repetition inspection tasks. In a real-world deployment at a converter station in Zhejiang Province, DEEP Robotics' quadruped robots have already achieved an average of over 1,000 hours of fault-free operation. The entire inspection process is managed by the self-developed DEEP Robotics Intelligent Patrol System, which provides full-cycle intelligent management and control. Utilizing intelligent recognition and self-calibration technologies, the system maintains a verified long-term recognition accuracy of over 95%. In complex environments, it can proactively respond to unexpected situations - ensuring autonomous inspection and return-to-base even in the event of network disconnection. A single system can manage multiple robot dogs simultaneously, significantly improving inspection efficiency and coverage across large-scale sites. This DEEP Robotics Intelligent Patrol System is set for official release and will accelerate the large-scale adoption of intelligent inspections by quadruped robots. Full Product Lineup Unveiled, Showcasing Breakthroughs in Embodied Intelligence Capabilities In the on-site product demonstration area, DEEP Robotics presented its comprehensive lineup of quadruped robots and industry-tailored application solutions, highlighting the latest advancements in key embodied intelligence technologies such as intelligent perception and interaction, as well as motion control. Among them, Lite3 stood out as the centerpiece of the product experience zone, becoming a popular highlight for interactive engagement with visitors. In collaboration with Daily Interactive, DEEP Robotics introduced an innovative experience where attendees could interact with Lite3 in a groundbreaking way: by wearing AR glasses, users could seamlessly switch to the robot dog's first-person perspective and, without the need for a handheld controller, intuitively command the robot using gesture-based controls enabled by precise motion recognition. This immersive interaction vividly demonstrated the promising future of integrating embodied intelligence robots with large-scale AI models. Also drawing significant attention at the exhibition were the LYNX wheeled-legged robot and its industrial variant, LYNX M20 - both of which previously went viral on domestic and international social media, sparking widespread interest and lively discussion. At WAIC 2025, the robots wowed the audience by executing a series of high-difficulty maneuvers such as continuously climbing 70cm platforms, performing "backflips into handstands," and executing "Thomas spins," showcasing their exceptional motion control and terrain adaptability. The flagship quadruped robot X30 was also featured alongside a range of industry-specific solutions - including intelligent power inspection, security patrol, and emergency firefighting - demonstrating DEEP Robotics' full-stack technical capabilities and scenario-driven delivery strength in deeply integrating its solutions into various industry needs. As one of the world's most influential AI events, WAIC has provided DEEP Robotics with an exceptional platform to showcase the strength of China's embodied intelligence technologies. To date, DEEP Robotics' products and solutions have been successfully implemented in over 600 industry scenario projects across 44 countries and regions globally and 34 provincial-level administrative regions in China, covering fields such as construction surveying, industrial operations and maintenance, emergency firefighting, power inspection, security patrol, and educational research. Looking ahead, DEEP Robotics will continue to build on independent innovation, increasing its R&D investment to drive ongoing breakthroughs in core technologies. At the same time, driven by practical implementation, the company will continue to explore the unique demands of different industries, leading the way in the deep application of embodied intelligence robots across a growing range of sectors. DEEP Robotics is committed to advancing the large-scale adoption of embodied intelligence and contributing to a more efficient, safer, and smarter future for industries worldwide. Company: DEEP RoboticsContact: Vera HuangEmail: huanglingxiao@ Form: SOURCE: DEEP Robotics View the original press release on ACCESS Newswire Error in retrieving data Sign in to access your portfolio Error in retrieving data Error in retrieving data Error in retrieving data Error in retrieving data
Associated Press
27-07-2025
- Business
- Associated Press
DEEP Robotics Showcases Core Technologies & Real-World Applications of Embodied Intelligence Robots at WAIC 2025
SHANGHAI, CN / ACCESS Newswire / July 27, 2025 / On July 26, the 2025 World Artificial Intelligence Conference (WAIC) and High-Level Meeting on Global AI Governance kicked off in grand fashion at the Shanghai World Expo Center. DEEP Robotics, a leading innovator in embodied intelligence technology and industrial applications, took center stage at the WAIC Expo Hall. For the first time, DEEP Robotics brought real-world operation scenarios of embodied intelligence into the exhibition, showcasing the full-process inspection capabilities of industrial-grade quadruped robots on site. The latest iterations of the Lite series robots, along with multiple industry application solutions, were unveiled simultaneously, demonstrating the company's leadership in both core technology breakthroughs and real-world implementations. This year's conference spans three days, with exhibition space exceeding 70,000 square meters for the first time, attracting over 800 leading enterprises both from home and abroad to Shanghai. Under the theme - Intelligent Connectivity for a Shared Future - the event presents a comprehensive ecosystem from AI core technologies to practical industry applications. Within its 150-square-meter immersive exhibition area, DEEP Robotics became a highlight of the event by recreating a real electric-power inspection scenario, performing live demonstrations of full-process operations, and showcasing the embodied intelligence capabilities of its newly iterated core products. Power Inspection Scenario Faithfully Recreated, Demonstrating Industrial-Grade Capabilities As a pioneer in industry-focused embodied intelligence solutions, DEEP Robotics built a dedicated industry demonstration zone at the exhibition, authentically recreating a power inspection scenario to vividly showcase the real-world operational capabilities of quadruped robots in industrial environments. In the simulated power inspection setup at the booth, X30 quadruped robot took on the role of an efficient inspector, accurately identifying inspection targets such as analog gauges, digital and status meters, infrared temperature readings, and indicator lights. To simulate the endurance demands of real industrial environments, the exhibition area was also equipped with a charging pile. X30 can autonomously monitor its battery level and navigate to the charging station when needed, enabling unattended continuous operation. This setup offered visitors a clear view of the feasibility of quadruped robots replacing manual labor in high-risk, high-repetition inspection tasks. In a real-world deployment at a converter station in Zhejiang Province, DEEP Robotics' quadruped robots have already achieved an average of over 1,000 hours of fault-free operation. The entire inspection process is managed by the self-developed DEEP Robotics Intelligent Patrol System, which provides full-cycle intelligent management and control. Utilizing intelligent recognition and self-calibration technologies, the system maintains a verified long-term recognition accuracy of over 95%. In complex environments, it can proactively respond to unexpected situations - ensuring autonomous inspection and return-to-base even in the event of network disconnection. A single system can manage multiple robot dogs simultaneously, significantly improving inspection efficiency and coverage across large-scale sites. This DEEP Robotics Intelligent Patrol System is set for official release and will accelerate the large-scale adoption of intelligent inspections by quadruped robots. Full Product Lineup Unveiled, Showcasing Breakthroughs in Embodied Intelligence Capabilities In the on-site product demonstration area, DEEP Robotics presented its comprehensive lineup of quadruped robots and industry-tailored application solutions, highlighting the latest advancements in key embodied intelligence technologies such as intelligent perception and interaction, as well as motion control. Among them, Lite3 stood out as the centerpiece of the product experience zone, becoming a popular highlight for interactive engagement with visitors. In collaboration with Daily Interactive, DEEP Robotics introduced an innovative experience where attendees could interact with Lite3 in a groundbreaking way: by wearing AR glasses, users could seamlessly switch to the robot dog's first-person perspective and, without the need for a handheld controller, intuitively command the robot using gesture-based controls enabled by precise motion recognition. This immersive interaction vividly demonstrated the promising future of integrating embodied intelligence robots with large-scale AI models. Also drawing significant attention at the exhibition were the LYNX wheeled-legged robot and its industrial variant, LYNX M20 - both of which previously went viral on domestic and international social media, sparking widespread interest and lively discussion. At WAIC 2025, the robots wowed the audience by executing a series of high-difficulty maneuvers such as continuously climbing 70cm platforms, performing 'backflips into handstands,' and executing 'Thomas spins,' showcasing their exceptional motion control and terrain adaptability. The flagship quadruped robot X30 was also featured alongside a range of industry-specific solutions - including intelligent power inspection, security patrol, and emergency firefighting - demonstrating DEEP Robotics' full-stack technical capabilities and scenario-driven delivery strength in deeply integrating its solutions into various industry needs. As one of the world's most influential AI events, WAIC has provided DEEP Robotics with an exceptional platform to showcase the strength of China's embodied intelligence technologies. To date, DEEP Robotics' products and solutions have been successfully implemented in over 600 industry scenario projects across 44 countries and regions globally and 34 provincial-level administrative regions in China, covering fields such as construction surveying, industrial operations and maintenance, emergency firefighting, power inspection, security patrol, and educational research. Looking ahead, DEEP Robotics will continue to build on independent innovation, increasing its R&D investment to drive ongoing breakthroughs in core technologies. At the same time, driven by practical implementation, the company will continue to explore the unique demands of different industries, leading the way in the deep application of embodied intelligence robots across a growing range of sectors. DEEP Robotics is committed to advancing the large-scale adoption of embodied intelligence and contributing to a more efficient, safer, and smarter future for industries worldwide. Company: DEEP Robotics Contact: Vera Huang Email: [email protected] Inquiry Form: SOURCE: DEEP Robotics press release
Yahoo
17-07-2025
- Business
- Yahoo
IDC Spotlights SUPERWISE® in New Report: Only 7.9% of Enterprises Are Ready to Operate AI at Scale
Discover why agentic AI demands new governance strategies—and how SUPERWISE helps enterprises move from experimentation to scalable, compliant AI operations NASHVILLE, Tenn., July 15, 2025--(BUSINESS WIRE)--IDC has released a new Spotlight report, "Governing and Operating AI at Scale," sponsored by SUPERWISE, highlighting the growing urgency for enterprises to establish strong governance foundations as agentic AI rapidly expands across business-critical operations. Written by IDC Research Director Shannon Kalvar, the report explores how enterprises are transitioning from traditional AI models to probabilistic, autonomous agentic AI—systems that perceive, reason, and act in pursuit of objectives. According to IDC's research, while 68% of companies are investing in agentic AI, only 7.9% are mature enough to operate and govern these systems at scale. "Agentic AI marks the most profound computing shift since mainframes," says Kalvar. "Enterprises that fail to govern it risk operational failures, compliance gaps, and eroding customer trust." Key Challenges Identified: Rising brand and regulatory risks from hallucinations and toxicity Difficulty managing hybrid work and aspatial, asynchronous AI operations Immature governance strategies in a fast-moving agentic landscape The report calls for a unified approach to AI governance and observability, especially as AI agents increasingly replace deterministic workflows with adaptive, semiautonomous systems. Why Download the Report: Understand why AI governance is now existential—not just strategic Learn how leading companies are establishing agent guardrails and oversight See how Superwise's unified platform addresses compliance, agent lifecycle, and observability Get actionable insights from IDC's maturity benchmarks and field data "Enterprises don't just need better models—they need systems that can be governed," said Russ Blattner, CEO of SUPERWISE "This IDC report validates what we're hearing from the field: AI governance isn't a checkbox—it's a control center for the future." About SUPERWISESUPERWISE is the Enterprise AI Governance and Operations Platform purpose-built for real-world AI. With a platform-first approach, it unifies artificial intelligence operations and governance, risk, and compliance into a single, scalable foundation, enabling complex industries to deploy, run, monitor, and manage AI with confidence and proper oversight. Recognized by Gartner as a Cool Vendor in Enterprise AI Governance and a pioneer in MLOps, SUPERWISE® delivers built-in guardrails, observability, explainability, and compliance, positioning AI as a trusted foundation for successful business transformation. For more information about SUPERWISE and its XAI capabilities, Download the Full IDC Spotlight Report. View source version on Contacts Media Contact:Jo Ann M StadtmuellerSVP, Marketingp: (615) 619-4033e: Error in retrieving data Sign in to access your portfolio Error in retrieving data Error in retrieving data Error in retrieving data Error in retrieving data
Forbes
14-07-2025
- Business
- Forbes
McDonald's ‘123456' Password Scare Reframes Responsible AI Debate
A security flaw on the McHire platform jeopardized 64 million applicants' data. Set aside aspirational AI rhetoric, alarmist consultant pitches and techno-babble. AI success requires candor about incentives, incompetence and indifference. McDonald's learned that harsh lesson (in a relatively costless way) when two security researchers used '123456' as the username and password to astonishingly fully access the Golden Arches hiring platform — and over 64 million applicants' personal data. The noble cyber sleuths, Ian Carroll and Sam Curry, reported the flaw to McDonald's and its AI vendor, Paradox, for swift technical resolution. If nefarious actors found the lax vulnerability, McDonald's leadership would be mired in a costly, public crisis. So, will the fast-food goliath learn from this 'near-miss' to improve tech governance? Will others tap this averted disaster for overdue responsible AI introspection and action? It depends. Widespread and hushed AI deployment problems need thornier fixes than many boards and senior executives will acknowledge, admit or address. Super-sized opportunities Workplace crises can be proactively prevented (or eventually explained) by tackling incentives, incompetence and indifference with stewardship, capability and care. The Golden Arches 'near miss' exemplifies that and the timing couldn't be better. While 88% of executives surveyed by PwC expect agentic AI spending increases this year, many struggle to articulate how AI will drive competitive advantage. Nearly 70% indicated that still half or fewer of their workforce interacts with agents daily. Indiscriminately 'throwing money' at issues can create more problems than it solves. Here's a better start. Dissect incentives. Talent, culture and bureaucratic entrenchment stymie big firms desperate to innovate. Nimble, bootstrapped startups tantalizingly fill those voids, but crave revenue and reputation. Stalled AI implementations only fuel that magnetism. Typically, the larger organization the makes headlines when deals falter. How many leadership teams meaningfully assess third-party risk from an incentives perspective? Or do expedited results more strongly appeal to their own compensation and prestige hunger? Is anyone seriously assessing which party has more (or less) to lose? Nearly 95% of McDonald's 43,000 restaurants are franchised. With over 2 million workers and aggressive growth aims, automating job applications is a logical AI efficiency move. Its selected vendor, whose tagline boasts 'meet the AI assistant for all things hiring' seemed like a natural partner. At what hidden costs? Successful strategic alliances require an 'outside-in' look at a counterparty's interests. Three of the seven-member Paradox board are private equity partners, including chair Mike Gregoire. In Startups Declassified, acclaimed business school professor and tech thought leader Steve Andriole emphasizes flagship revenue's valuation criticality, 'There's no more important start-up activity than sales — especially important are the 'lighthouse' customers willing to testify to the power and greatness of products and services. Logo power is [vital] to start-ups.' 'Remember that no one wants to buy start-ups unless the company has killer intellectual property or lists of recurring customers. Profitable recurring revenue is nirvana. Exits occur when a start-up becomes empirically successful,' he continued. Assess skill and will. Despite its global presence, digital strategy imperatives and daily transaction volume, the 2025 McDonald's proxy reveals three common AI-era oversight shortfalls: inadequate boardroom cyber expertise, no technology committee and cybersecurity relegated to audit oversight. Those are serious signaling problems. In fact, the word 'cybersecurity' only appears nine times across the 100-page filing. In the director qualifications section, information technology is grouped with cybersecurity and vaguely defined 'contributes to an understanding of information technology capabilities, cloud computing, scalable data analytics and risks associated with cybersecurity matters.' Just four of the eleven directors are tagged as such. While three of those four worked in the tech sector, none has any credible IT or cybersecurity expertise. Intriguingly, not one of the four, board member and former Deloitte CEO Cathy Engelbert has the best experience to push stronger governance. Is she, now the prominent WNBA league commissioner, willing to take such contentious risk? To start, she can tap longtime McDonald's CFO Ian Borden and auditors EY for guidance and ideas on bolstering board composition. Nearly 95% of McDonald's 43,000 restaurants worldwide are franchised. When tech issues arise, fingers, by default, point at the IT team. However, responsible AI design and deployment truly require cross-functional leadership commitment. McDonald's CEO Chris Kempczinski routinely touts a 4D strategy (digital, delivery, drive-thru and development) and characterizes the fast-food frontrunner's tech edge as 'unmatched.' That bravado brings massive expectations and he can't be happy with the '123456' password distraction. With annual compensation approaching $20 million annually, he also has a responsible AI obligation to current and future McDonald's workers making, on average, 1,014 times less — as well as the 40,000 franchisees. Valerie Ashbaugh, McDonald's commercial products and platform SVP, rotates into the US CIO seat next month. The timing is ideal to institute policies, procedures and accountability for stronger third-party IT access controls. Alan Robertson, UK ambassador to the Global Council for Responsible AI, astutely notes, 'The damage is done — not by hackers, but by sheer negligence. McDonald's has pinned the issue on Paradox. Paradox says they fixed it and have since launched a bug bounty program. It raises bigger questions for all of us. Who audits the third-party vendors we automate hiring with? Where does the liability sit when trust is breached at this scale? And what does 'responsible AI' even mean when basic cybersecurity hygiene isn't in place? We talk about ethics — but sometimes it's just about setting a password.' That's prototypical indifference — especially when the access key is "123456." Likewise, HR leaders have a chance to meaningfully shape AI rollouts. 'HR needs to resist the urge to 'just go along.' There will be many HR leaders who simply wait for the various software lines they current license to add AI functionality. To do so would be a mistake. AI will become a critical part of the employee experience and HR should have a hand in that,' advises AthenaOnline SVP of customer solutions Mark Jesty. At McDonald's, EVP and global chief people officer Tiffanie Boyd holds that golden opportunity to elevate responsible AI on the board and c-suite agendas. Will she? Responsibility knocks The McHire 'near-miss' highlights how boards and c-suites can remain dangerously unprepared for AI design, deployment and oversight. Strategy speed and tech wizardry must never be at stewardship's cost. "If you're deploying AI without basic security hygiene, you're not innovating. You're endangering people. Security is not optional,' implores CEO Ivan Rahman. Who's opting for drive-thru AI governance?
