[AI新闻] 2026年欧洲AI工程师大会[AINews] AI Engineer Europe 2026

1. Claude Emerges as Preferred AI Chatbot at HumanX Conference
   At the HumanX AI conference in San Francisco, attendees and vendors consistently cited Anthropic’s Claude as the most popular AI chatbot, particularly for enterprise and consumer applications. Discussions across panels and vendor interactions highlighted a growing preference for Claude over OpenAI’s ChatGPT, with several participants expressing dissatisfaction with ChatGPT’s recent direction. One vendor explicitly stated that his team relied heavily on Claude, claiming ChatGPT had “fallen off” in performance and reliability.

This sentiment reflects a broader shift in perception within the AI community. Despite OpenAI’s recent $122 billion funding round and anticipated IPO, industry observers note a decline in confidence regarding the company’s strategic focus and leadership. Multiple attendees questioned OpenAI’s ability to maintain its early momentum amid internal and external challenges.

Key Takeaways:
Claude gains traction as top choice among AI professionals at major conference
OpenAI faces growing skepticism over strategic direction and product quality
Industry professionals report declining satisfaction with ChatGPT compared to Claude

Source: Original Article

1. OpenAI Enhances Sales Efficiency with ChatGPT Integration
   ChatGPT streamlines sales workflows by automating time-consuming tasks such as research, preparation, follow-up, and deal coordination. It processes unstructured inputs like CRM data, call notes, and account information to generate structured outputs including briefs, emails, and action plans. This allows sales teams to focus more on customer engagement and maintain consistency across outreach and execution.

The tool accelerates account and meeting preparation by synthesizing internal context, identifying knowledge gaps, and producing concise pre-call briefs and follow-up strategies. It reduces manual effort while ensuring no critical details are overlooked before customer interactions.

ChatGPT improves outreach personalization and consistency by drafting tailored emails, call recaps, and next-step messages aligned with team tone and structure. It supports multi-channel sequences across email and LinkedIn, enabling scalable yet individualized communication.

Key Takeaways:
ChatGPT automates sales prep and follow-up tasks
Tool generates structured outputs from unstructured sales data
Improves consistency and personalization in customer outreach

Source: Original Article

1. OpenAI Launches ChatGPT for General and Enterprise Use
   ChatGPT serves as OpenAI’s primary user-facing product, offering a conversational interface for writing, learning, problem-solving, and creative tasks. It supports broad applications such as summarization, tutoring, brainstorming, and planning. Designed for general-purpose use, it is accessible to individuals and professionals across various domains.

The platform also includes specialized versions like ChatGPT Enterprise and ChatGPT Business, tailored for organizational needs. These versions provide enhanced administrative controls, improved privacy, team collaboration tools, and deployment support. They enable businesses to integrate AI securely into workflows while maintaining data governance and compliance standards.

ChatGPT reflects OpenAI’s shift from research-focused development to real-world AI deployment. By offering both consumer and enterprise solutions, OpenAI aims to make advanced AI accessible while addressing scalability and security concerns. This dual approach supports widespread adoption across education, business, and creative industries.

Key Takeaways:
ChatGPT enables diverse tasks through conversational AI
Enterprise versions offer enhanced security and collaboration
OpenAI transitions from research to practical AI deployment

Source: Original Article

1. OpenAI Introduces Codex for AI-Assisted Software Development
   Codex is OpenAI’s specialized tool designed to assist developers with coding tasks such as writing, debugging, refactoring, and navigating programming workflows. It operates within IDEs or development environments, providing context-aware suggestions and automation. Unlike general chat interfaces, Codex focuses on technical accuracy and integration with software development processes.

By understanding natural language prompts, Codex translates user intent into functional code across multiple programming languages. It supports rapid prototyping, reduces repetitive coding tasks, and helps developers learn new frameworks. This increases productivity and lowers barriers to software creation for both novice and experienced programmers.

Codex exemplifies OpenAI’s strategy of targeting specific professional domains with tailored AI tools. Its integration into development environments highlights the growing role of AI in augmenting human expertise. The tool contributes to broader trends in AI-driven automation within the tech industry.

Key Takeaways:
Codex automates coding tasks within development environments
It supports multiple languages and improves developer productivity
OpenAI targets specialized professional use cases with AI

Source: Original Article

1. OpenAI API Enables Developers to Integrate AI Models
   The OpenAI API provides developers with programmable access to OpenAI’s suite of models for text, image, and code generation. It allows integration of AI capabilities into custom applications, workflows, and systems. Developers can build tools for content analysis, reasoning, tool interaction, and multimodal tasks.

This API supports scalable deployment of AI features without requiring in-house model training. It enables startups and enterprises to leverage advanced AI while focusing on application logic and user experience. The composable nature of the API encourages innovation across industries, from education to healthcare.

By offering flexible access, OpenAI extends its impact beyond standalone products. The API represents a key component of OpenAI’s ecosystem, fostering third-party development and broader AI adoption. It underscores the importance of developer platforms in democratizing advanced technology.

Key Takeaways:
OpenAI API allows integration of AI into custom applications
It supports text image and code generation tasks
Developer access promotes innovation and scalable AI deployment

Source: Original Article

1
ChatGPT Enhances Finance Team Efficiency Through Structured Workflows

Finance teams often face challenges in transforming incomplete data into reliable reports due to time-consuming tasks like reconciling numbers, explaining variances, and updating forecasts. ChatGPT reduces this overhead by organizing unstructured inputs, drafting initial outputs, and standardizing recurring workflows. It supports finance professionals by streamlining preparation and improving consistency without replacing human judgment.

The tool helps structure complex problems by outlining key questions, identifying drivers to analyze, and suggesting follow-up actions. This enables teams to begin work more efficiently while maintaining thoroughness. By clarifying dense financial communication, ChatGPT makes updates more accessible to non-finance stakeholders while preserving accuracy and necessary caveats.

Key Takeaways:
ChatGPT reduces time spent on formatting and rewriting in finance workflows
Tool standardizes recurring deliverables like variance commentary and forecasts
Improves clarity of financial communication for non-expert audiences

Source: Original Article

2
ChatGPT Supports Forecasting and Planning with Structured Frameworks

In forecasting and planning, finance teams use ChatGPT to build models, assess scenarios, and manage budgets and headcount. The AI generates assumption checklists, driver frameworks, and scenario tables to support structured decision-making. It also identifies critical questions to validate forecasts, enhancing the rigor of financial planning processes.

By providing consistent templates and logical structures, ChatGPT helps teams avoid rebuilding models from scratch each cycle. This increases efficiency and ensures reviewers can quickly locate key information. The tool acts as a collaborative assistant, accelerating preparation while maintaining analytical integrity.

Key Takeaways:
ChatGPT generates assumption checklists and scenario tables for forecasting
Tool improves consistency in financial planning deliverables
Supports validation of forecasts through structured questioning

Source: Original Article

1. Treasury and Fed Urge Banks to Test Anthropic’s Mythos Model for Security
   U.S. Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell convened a meeting with major bank executives this week, encouraging them to adopt Anthropic’s newly released Mythos AI model to identify cybersecurity vulnerabilities. The model, not specifically trained for security purposes, has demonstrated a strong ability to detect system weaknesses, prompting interest from financial institutions.

JPMorgan Chase was the only bank officially listed as an initial partner with access to Mythos, but Goldman Sachs, Citigroup, Bank of America, and Morgan Stanley are reportedly conducting internal tests. Anthropic has limited access to the model due to concerns about its effectiveness in uncovering security flaws, raising questions about responsible deployment.

The recommendation is notable given Anthropic’s ongoing legal dispute with the Trump administration over a Department of Defense designation labeling the company a supply-chain risk. That classification followed failed negotiations over restrictions on government use of Anthropic’s AI systems.

Key Takeaways:
U.S. financial regulators promote AI for cybersecurity despite access limitations
Major U.S. banks testing Mythos despite Anthropic’s restricted rollout
Anthropic faces government scrutiny while gaining financial sector interest

Source: Original Article

1. U.K. Regulators Assess Risks of Anthropic’s Mythos Model in Finance
   U.K. financial regulators are engaging in discussions about the potential risks associated with Anthropic’s Mythos AI model, according to the Financial Times. The model’s ability to detect security vulnerabilities has drawn attention from global financial authorities, prompting evaluations of its implications for systemic risk and data protection.

Although no formal restrictions have been announced, regulators are examining how widespread use of such AI tools could affect financial stability and compliance frameworks. The discussions reflect growing concern over advanced AI systems operating in critical infrastructure without standardized oversight.

This scrutiny aligns with broader international efforts to manage AI deployment in sensitive sectors. The U.K.’s proactive stance may influence future regulatory approaches in other jurisdictions.

Key Takeaways:
U.K. regulators evaluating Mythos for financial sector risks
Concerns focus on security and systemic implications of AI tools
Global regulatory coordination on AI in finance may accelerate

Source: Original Article

1
Clinicians Use ChatGPT for Diagnosis and Patient Care with HIPAA-Compliant Tools

Healthcare professionals are increasingly integrating ChatGPT into clinical workflows to assist with diagnosis, medical documentation, and patient communication. OpenAI has developed HIPAA-compliant versions of its AI tools, enabling secure use in medical settings. These tools help reduce administrative burden by generating clinical notes, summarizing patient histories, and suggesting diagnostic considerations based on input data.

The adoption reflects a broader trend toward AI-assisted healthcare, where clinicians use AI to enhance efficiency without replacing human judgment. Institutions using these tools report improved documentation speed and reduced burnout among providers. However, oversight remains critical to ensure accuracy and compliance with medical standards.

Key Takeaways:
Clinicians leverage ChatGPT for diagnosis and documentation tasks
HIPAA-compliant AI tools enable secure healthcare applications
AI supports efficiency but requires human oversight for accuracy

Source: Original Article

1. OpenAI Rotates Certificates and Updates Apps After Axios Supply Chain Attack

OpenAI responded to the recent Axios supply chain attack by rotating its macOS code signing certificates and releasing updated versions of its applications. The company confirmed that no user data was compromised during the incident. This proactive measure aimed to prevent further exploitation of potentially affected systems and maintain trust in its software distribution channels.

The attack targeted Axios, a third-party developer tool used by OpenAI, which may have allowed malicious actors to inject unauthorized code into software builds. Although OpenAI stated its systems were not directly breached, the potential risk prompted immediate remediation steps. Certificate rotation ensures that only newly signed and verified applications can run on macOS devices.

Key Takeaways:
OpenAI rotated macOS code signing certificates post-Axios attack
No user data was compromised during the supply chain incident
Updated apps released to prevent potential malicious code execution

Source: Original Article

1. Industry Leaders Advocate for Open Model Consortium to Sustain Frontier AI Development

The push for a consortium to fund foundational open AI models is gaining momentum, driven by concerns over sustainability and access. Stanford professor Percy Liang, leader of the fully open Marin project, highlights the inevitability of such a collaborative effort. While Nvidia’s Nemotron initiative represents a single-company attempt to support open models, long-term stability likely requires multi-entity funding.

Recent instability in open model labs—marked by high-profile exits at Qwen and Ai2—underscores the financial pressures facing independent developers. These trends mirror Meta’s earlier pivot away from Llama, signaling a broader industry shift. As training costs rise, maintaining frontier-level open models becomes increasingly difficult for individual organizations.

Chinese startups like Moonshot AI, MiniMax, and Z.ai currently contribute strong open models but face uncertain futures due to funding constraints. Their ability to sustain R&D investment remains in question. Open releases often conflict with revenue-generating AI product development, creating strategic tension for companies balancing innovation and profitability.

Key Takeaways:
Consortium model seen as only sustainable path for frontier open AI
Rising training costs challenge independent open model labs
Revenue pressures push firms toward closed or limited open releases

Source: Original Article

1. Apple Develops Four Smart Glasses Designs for 2027 Launch
   Apple is testing four distinct smart glasses designs ahead of a planned 2027 release, with a potential unveiling as early as late 2024, according to Bloomberg’s Mark Gurman. The designs include a large rectangular frame, a slimmer rectangular style resembling Tim Cook’s glasses, a larger oval or circular frame, and a smaller oval or circular version. Multiple color options such as black, ocean blue, and light brown are also under consideration.

The smart glasses represent a scaled-back vision compared to Apple’s earlier ambitions for mixed and augmented reality devices, which faced setbacks including delays and the underwhelming market response to the Vision Pro. Unlike the Vision Pro, these glasses will not feature displays but will support photo and video capture via oval camera lenses, phone calls, music playback, and integration with an upgraded Siri.

This move positions Apple closer to Meta’s Ray-Ban smart glasses, focusing on audio and camera functionality rather than immersive visuals. The shift reflects a strategic pivot toward wearable tech with broader consumer appeal and lower technical complexity.

Key Takeaways:
Apple targets 2027 for smart glasses release with multiple design options
Devices will lack displays but support camera audio and Siri functions
Strategy marks a retreat from complex AR to simpler wearable tech

Source: Original Article

1. Trump Asserts US Energy Independence Amid Iran Conflict and Rising Gas Prices

President Donald Trump claimed the United States is energy-independent and unaffected by disruptions in global oil markets caused by the conflict with Iran. Speaking last week, he emphasized that the U.S., as the world’s largest oil and gas producer, does not rely on tankers passing through the Strait of Hormuz, which Iran blocked for a month. Despite his assurances, domestic gasoline prices have surged above $4 per gallon for the first time in four years.

U.S. households paid an additional $8.4 billion for gasoline over the past month compared to pre-conflict levels, according to a report by Democrats on Congress’ Joint Economic Committee. Although a two-week ceasefire agreement was reached, most tankers remained blocked as both sides disputed terms. Iran continues to assert control over the Strait of Hormuz, through which 20% of global oil and LNG passes, and has begun charging multimillion-dollar tolls for passage.

The U.S. Energy Information Administration forecasts elevated oil prices through year-end, even if the conflict resolves by April. Key infrastructure, including Qatar’s Ras Laffan LNG terminal, sustained missile damage that could impair operations for years. The global energy shock has led to gas rationing in parts of Asia, flight cancellations in Europe, and widespread fuel shortages.

Key Takeaways:
Trump claims US energy independence despite rising gas prices
Households paid $8.4 billion more for gasoline amid Iran conflict
Global energy markets face prolonged disruption from Strait of Hormuz tensions

Source: Original Article

1. NASA Orion Spacecraft Completes Historic Moon Mission with Safe Pacific Splashdown
   NASA’s Orion spacecraft, named Integrity, successfully returned to Earth on Friday after completing humanity’s first crewed lunar voyage in nearly 54 years. Traveling at over 30 times the speed of sound, the capsule reentered Earth’s atmosphere, enduring external temperatures reaching approximately 5,000 degrees Fahrenheit. A plasma sheath temporarily disrupted communications with Mission Control in Houston during a six-minute blackout period.

The spacecraft executed a controlled descent southwest of San Diego, deploying a series of parachutes to stabilize and slow its plunge. Three main chutes, each covering 10,500 square feet, opened to reduce velocity before splashdown at 8:07 pm EDT. In just 14 minutes, Orion decelerated from nearly 25,000 mph, subjecting the crew to two brief periods of 3.9 Gs.

Recovery operations were conducted by the USS John P. Murtha, which deployed helicopters and small boats to retrieve the four astronauts. Commander Reid Wiseman confirmed all crew members were in good health. Christina Koch exited first, followed by Victor Glover and Canadian astronaut Jeremy Hansen. Wiseman was the last to leave the capsule.

Key Takeaways:
NASA successfully completes first crewed lunar mission in over five decades
Orion spacecraft endures extreme reentry conditions and lands safely in Pacific Ocean
Four astronauts including first woman and person of color on lunar mission return unharmed
US Navy and NASA coordinate complex recovery operation following historic splashdown

Source: Original Article

1. NASA Completes Artemis II Mission with Pacific Ocean Splashdown

NASA successfully concluded the Artemis II mission on Friday evening as its Orion spacecraft, carrying four astronauts, splashed down in the Pacific Ocean off California after a 700,000-mile journey around the Moon. This marked humanity’s first crewed deep space mission in over 50 years and a major milestone for NASA and its international partners. The mission demonstrated key capabilities of the Orion spacecraft and the Space Launch System rocket.

The successful return validates critical systems for future lunar missions, including life support, navigation, and re-entry procedures. NASA officials praised the mission’s precision, particularly the rocket’s orbital accuracy exceeding 99%. However, the agency emphasized that Artemis II was a foundational step, with more complex challenges ahead for sustained lunar exploration.

Key Takeaways:
Artemis II completes first crewed lunar flyby in over 50 years
Mission validates Orion spacecraft and SLS rocket performance
NASA prepares for more complex Artemis III and IV missions

Source: Original Article

1. NASA Advances Space Launch System for Artemis III Mission

NASA confirmed the core stage for the Artemis III mission will depart Michoud Assembly Facility in Louisiana later this month for delivery to Kennedy Space Center in Florida. Other components of the Space Launch System rocket have already arrived or are en route. The SLS performed with high precision during Artemis II, achieving target orbit with over 99% accuracy.

Despite the rocket’s success, the Mobile Launch Tower sustained moderate damage during the Artemis II launch and will be moved to the Vehicle Assembly Building for repairs and refurbishment. This maintenance is essential for supporting upcoming launches. NASA continues to prepare infrastructure for increasingly complex missions involving multiple spacecraft and lunar landings.

Key Takeaways:
Artemis III SLS core stage nearing transport to launch site
Mobile Launch Tower requires refurbishment after Artemis II
NASA builds momentum for crewed lunar landing missions

Source: Original Article

1. Californians Sue Sutter Health and MemorialCare Over AI Recording Without Consent
   A group of California patients filed a federal class-action lawsuit against Sutter Health and MemorialCare, alleging that an AI transcription tool recorded their medical consultations without informed consent. The lawsuit, filed in San Francisco, claims the AI system captured sensitive physician-patient conversations during recent medical visits. Plaintiffs assert they were not informed that their private health discussions would be processed by artificial intelligence or transmitted outside clinical environments.

The complaint identifies Abridge AI as the tool used to record and process confidential medical communications. These recordings included personal health information such as symptoms, diagnoses, medications, and treatment plans. The plaintiffs argue this practice violates both state and federal privacy laws, including California’s strict confidentiality regulations for medical data.

Abridge’s AI technology has seen rapid adoption across major U.S. healthcare systems, including Kaiser Permanente, Mayo Clinic, and Duke Health. The lawsuit raises broader concerns about patient privacy and the transparency of AI deployment in clinical settings. Legal experts suggest the case could set a precedent for how AI tools are regulated in healthcare.

Key Takeaways:
Patients allege AI recorded medical visits without consent
Abridge AI used across major healthcare providers nationwide
Lawsuit challenges legality of AI in confidential patient interactions

Source: Original Article

1. Kepler Communications Launches Largest Orbital Compute Cluster with Nvidia Orin Processors

Kepler Communications, a Canadian satellite company, has deployed the largest compute cluster currently in orbit, featuring 40 Nvidia Orin edge processors across 10 operational satellites linked via laser communications. The system, launched in January, supports real-time data processing in space and currently serves 18 customers. This infrastructure enables faster analysis of sensor data collected in orbit, reducing reliance on ground-based systems.

The company positions itself not as a traditional data center provider but as a space-based network infrastructure layer. It aims to support other satellites, drones, and aircraft by offering distributed computing and communication services. This model could become foundational for future space-based applications requiring low-latency data handling.

Key Takeaways:
Kepler operates the largest orbital compute cluster using Nvidia Orin chips
The system processes data in space to enhance satellite sensor performance
Kepler provides infrastructure services for satellites and aerial platforms

Source: Original Article

1. Sophia Space Partners with Kepler to Test Orbital Computer Software

Sophia Space, a startup developing passively-cooled space computers, has partnered with Kepler Communications to test its proprietary operating system in orbit. The software will be uploaded to one of Kepler’s satellites and configured across six GPUs on two spacecraft. This marks the first attempt to deploy and manage complex software across multiple orbital GPUs.

The collaboration serves as a critical validation step for Sophia ahead of its planned satellite launch in late 2027. Passive cooling technology could address a major challenge for orbital data centers—managing heat from high-performance processors without heavy active-cooling systems. Success in this test would demonstrate feasibility for scalable, cost-effective computing in space.

Key Takeaways:
Sophia Space will test its OS on Kepler’s satellite network
The experiment represents the first multi-GPU software deployment in orbit
Passive cooling innovation may enable future large-scale space data centers

Source: Original Article

Intel Secures Long Term AI Cloud Partnership with Google

Intel has entered into a long term strategic partnership with Google Cloud to integrate its AI hardware and software technologies into Google’s cloud infrastructure. The collaboration focuses on optimizing Intel’s Gaudi AI accelerators and Habana software stack for deployment across Google Cloud’s data centers. This integration aims to provide enterprises with scalable and cost effective AI training and inference solutions using Intel’s silicon.

The deal positions Intel as a stronger competitor in the AI chip market, currently dominated by Nvidia. By aligning with Google Cloud, Intel gains access to a major cloud platform with global reach and enterprise customers. The partnership is expected to drive adoption of Intel’s AI accelerators in hybrid and multi cloud environments, enhancing its relevance in the rapidly growing AI infrastructure sector.

Key Takeaways:
Intel strengthens AI cloud presence through Google partnership
Gaudi accelerators gain access to Google Cloud infrastructure
Deal challenges Nvidia’s dominance in AI hardware market

Source: Original Article