Chinese tech giant ByteDance is significantly expanding its artificial intelligence division, Seed, in the United States, hiring nearly 100 people. This strategic move aims to bolster its competitive stance against leading US AI firms, focusing on developing large language models and enhancing tools for text, image, and video generation, alongside science models for drug discovery. This expansion follows ByteDance's efforts to address US national security concerns by selling parts of its TikTok business. However, the company faces accusations from Hollywood regarding intellectual property theft, with major studios sending cease-and-desist letters over models like Seedance, which are noted for their speed and lower cost but also for fewer copyright safeguards.
While companies like ByteDance push AI boundaries, the technology presents both opportunities and challenges across various sectors. Rollick, for instance, has launched an AI Assistant to revolutionize online car shopping, integrating it into OEM websites to help customers explore products, ask questions, and find local dealer inventory. This assistant also pre-qualifies shoppers for financing, leveraging Rollick's proprietary data. On the corporate training front, Lockheed Martin successfully implemented its AI Roadshow program, standardizing AI use across its communications function. This initiative achieved 100% employee participation and measurable growth in AI adoption, with the company tracking usage and aiming for further increases by the end of 2026.
However, the rapid integration of AI also brings significant risks. Researchers found that using large language models like Claude, ChatGPT, and Gemini for password generation can lead to weak security, as these AIs struggle with true randomization, producing passwords with lower entropy that are easier to crack. Furthermore, cyber attackers are exploiting vulnerabilities in Customer Experience (CX) platforms, breaching nearly 700 organizations by "poisoning" unstructured data that feeds AI engines. These platforms, often overlooked by security teams, are critical systems connected to HR, CRM, and payment systems, highlighting a need for better data loss prevention and bot mitigation.
In higher education, the growing use of AI raises concerns beyond just cheating, potentially eroding the core of learning itself by weakening the mentorship ecosystem and promoting cognitive offloading. Misericordia University faculty are actively researching these impacts, presenting on ethical AI integration and human-centered design at a conference in Italy. Meanwhile, individuals like retired lawyer Marcel Strigberger are cautiously exploring AI's capabilities, noting its quick answer provision but also emphasizing the importance of critical thinking. Even in sports, Microsoft Copilot AI demonstrates its predictive power, generating a mock draft for the 2026 NBA first round based on team records and prospect analysis.
Key Takeaways
- ByteDance is expanding its US AI division, Seed, hiring nearly 100 people to develop large language models and AI tools, despite ongoing US security concerns.
- Chinese AI models, including ByteDance's Seedance, face accusations of intellectual property theft from Hollywood due to fewer copyright safeguards and competitive pricing.
- Rollick launched an AI Assistant for car shoppers, integrating into OEM websites to facilitate product discovery, answer questions, and pre-qualify financing.
- Lockheed Martin implemented an AI Roadshow training program, achieving 100% employee participation and measurable growth in AI adoption across its communications function.
- Using large language models like Claude, ChatGPT, and Gemini for password generation creates weak security due to their struggle with true randomization and lower entropy.
- Cyber attackers are exploiting vulnerabilities in Customer Experience (CX) platforms by "poisoning" unstructured data, leading to breaches in nearly 700 organizations.
- The widespread use of AI in higher education poses a risk of eroding core learning and mentorship, raising concerns about cognitive offloading and the purpose of universities.
- Misericordia University faculty are researching AI's impact on educator preparation and ethical integration in higher education, presenting at IHSI 2026 in Italy.
- Microsoft Copilot AI generated a mock draft for the 2026 NBA first round, predicting player selections based on current team records and prospect analysis.
- Individuals, like retired lawyer Marcel Strigberger, are cautiously exploring AI's capabilities for quick information retrieval while emphasizing critical thinking.
ByteDance expands US AI team amid security concerns
Chinese tech company ByteDance is hiring nearly 100 people in the US for its artificial intelligence division, Seed. This move aims to compete with top US AI firms despite ongoing national security worries from American lawmakers. The open roles involve developing large language models, improving AI tools for text, image, and video generation, and creating science models for drug discovery. ByteDance's expansion follows its deal to sell parts of its US TikTok business, intended to address security concerns about data collection and content manipulation. The company's AI chatbot Doubao was popular in China, and it recently launched new AI video and image models, facing accusations of intellectual property theft from Hollywood.
ByteDance boosts US AI presence with new jobs
ByteDance, a Chinese tech giant, is significantly expanding its artificial intelligence team in the United States by hiring for almost 100 positions. This expansion, focused on its AI division called Seed, aims to strengthen its competitive position against leading US AI companies. The new roles will contribute to developing large language models, advancing text, image, and video generation tools, and building science models for areas like drug discovery. This push comes after ByteDance addressed US national security concerns by selling parts of its TikTok business. The company's AI chatbot Doubao was a top performer in China, and its recent AI model launches have drawn criticism from Hollywood over intellectual property concerns.
Rollick launches AI assistant for car shoppers
Rollick has introduced a new AI Assistant designed to enhance the online car shopping experience for customers. This assistant will be integrated into OEM websites, allowing shoppers to explore products, ask questions, and find local dealer inventory through natural conversations. The AI Assistant aims to help manufacturers engage shoppers earlier in the buying process and can pre-qualify them for financing. It uses Rollick's proprietary data from customer interactions and ownership insights across various industries. Initially focused on product discovery and shopper engagement, the AI Assistant will later be available for dealer websites and Rollick's marketplaces.
AI risks eroding learning in higher education
The growing use of artificial intelligence in higher education poses a greater risk than cheating, potentially eroding the core of learning itself. While AI can assist with research and administrative tasks, its widespread adoption raises questions about the purpose of universities. As AI becomes more capable of producing knowledge work, it could weaken the learning and mentorship ecosystem. Concerns include transparency in AI interactions, accountability for AI-generated work, and the cognitive offloading that may prevent students from developing essential skills. Universities must consider how to integrate AI responsibly without diminishing the value of human learning and interaction.
Attackers exploit AI in CX platforms to breach organizations
Cyber attackers are exploiting vulnerabilities in Customer Experience (CX) platforms to breach nearly 700 organizations. These platforms process vast amounts of unstructured data, including survey responses and social media comments, which feed into AI engines. Attackers poison this data, causing the AI to perform harmful actions. Security teams often overlook these platforms, miscategorizing them as simple survey tools rather than critical systems connected to HR, CRM, and payment systems. Key security gaps include data loss prevention failures with unstructured data, live API tokens from old campaigns, lack of bot mitigation for public input channels, and lateral movement through approved API calls.
Lockheed Martin builds company-wide AI training program
Lockheed Martin has successfully implemented an AI training program called the AI Roadshow to standardize AI use across its communications function. The program surveyed employees to identify key use cases and tailored a learning funnel for different roles. This initiative resulted in 100% participation, clearer guidelines for AI use, and measurable growth in adoption. The training covers foundational AI knowledge, specific applications for text and visual communications, and advanced deep dives into specific use cases. Lockheed Martin is tracking AI usage, which has increased significantly, and aims for further growth by the end of 2026.
Misericordia University faculty share AI research in Italy
Faculty members from Misericordia University presented their research on artificial intelligence and human-centered design at the 9th International Conference on Human Intelligent Systems Integration (IHSI 2026) in Florence, Italy. The presentations focused on AI's impact on educator preparation, ethical AI integration in higher education, and using human-computer interaction principles in exercise science. One poster session explored assessing teacher candidates' professionalism through mixed reality. This participation highlights the university's commitment to advancing research at the intersection of technology and human development, emphasizing innovation, human dignity, and ethical responsibility.
Retired lawyer explores AI's capabilities
A retired lawyer, Marcel Strigberger, shares his hesitant journey into using artificial intelligence. Initially uncomfortable with technology, he decided to test AI after his family encouraged him. He asked basic questions like the capital of France and the meaning of the Miranda case, noting AI's ability to provide quick answers. Strigberger also observed that AI sessions seem to invite endless follow-up questions. While acknowledging AI's potential, he approaches it with caution, reflecting on its limitations and the importance of critical thinking, especially in the legal field.
AI password generators create weak security
Using large language models (LLMs) like Claude, ChatGPT, and Gemini to generate passwords can lead to weak security. Researchers found that these AI models struggle with true randomization, often repeating patterns and characters. Passwords generated by AI have significantly lower entropy, making them much easier for attackers to crack compared to truly random passwords. While individual users are advised against using AI for password creation, AI agents that rely on LLMs for tasks like coding may inadvertently create vulnerable passwords. This issue highlights a growing cybersecurity risk as AI becomes more integrated into various digital processes.
China's AI advancements challenge Hollywood's IP
China's rapid progress in artificial intelligence is creating significant challenges for Hollywood, particularly concerning intellectual property. Chinese AI models, like ByteDance's Seedance, are undercutting US rivals in price and speed, while also having fewer safeguards regarding copyrighted material. Major Hollywood studios and groups have sent cease-and-desist letters to ByteDance over Seedance, alleging copyright infringement and misuse of likenesses. Despite legal pressure, the broader rise of Chinese AI models continues, raising concerns about market dominance and the potential for China to shape global AI regulations. This situation represents an early battle in a larger conflict over intellectual property and market control.
AI predicts 2026 NBA mock draft first round picks
Microsoft Copilot AI has generated a mock draft for the 2026 NBA first round, predicting player selections based on current team records, needs, and prospect analysis. The AI assigned the draft order using current NBA standings, with the worst teams receiving the highest picks. The top three picks are anticipated to be highly sought after by NBA front offices, suggesting a strong class of college players approaching March Madness. The AI's predictions provide insights into potential team strategies and player trajectories for the upcoming draft.
Sources
- ByteDance building out artificial intelligence team in US
- ByteDance Building Out Artificial Intelligence Team in US
- Rollick Debuts AI Assistant Designed to Engage Shoppers
- The greatest risk of AI in higher education isn’t cheating – it’s the erosion of learning itself
- How attackers hit 700 organizations through CX platforms your SOC already approved
- How Lockheed Martin turned AI curiosity into a company-wide curriculum
- Misericordia University Faculty Present Research on Artificial Intelligence and Human-Centered Design at International Conference in Florence, Italy
- Old Attorney, New Tricks: 'Boomer lawyer' gives AI a try
- AI-Generated Passwords Are Apparently Quite Easy to Crack
- China's AI rise rattles Hollywood
- 2026 NBA mock draft by AI featuring first round picks, predictions
Comments
Please log in to post a comment.