intel launches anthropic while openai expands its platform

Key Takeaways

• Intel CEO Pat Gelsinger reports strong demand for AI accelerators and 14th Gen server CPUs, with PC market expected to decline but AI-powered PCs offsetting the drop.
• Harvard College now requires all freshmen in expository writing courses to complete a three-part AI training module covering LLMs, AI's impact on education, and societal questions.
• Zscaler joined Anthropic's Project Glasswing, using the Claude Mythos Preview AI model to find software vulnerabilities, and makes internal apps invisible to the internet via its Zero Trust Exchange platform.
• A CalMatters commentary criticizes AI companies like Anthropic, OpenAI, and Google DeepMind for military contracts and real-world harms, including a strike on a school in Iran that killed 18 children.
• OpenAI CEO Sam Altman apologized for not alerting law enforcement about the Tumbler Ridge shooter's use of ChatGPT, highlighting tension between user privacy and threat reporting.
• China's auto industry is embedding AI in cars following Beijing's five-year plan mandate, aiming to create self-reasoning electric vehicles running on Chinese chips and software.
• AI is helping doctors predict and detect kidney disease earlier using models like logistic regression, XGBoost, and deep neural networks, with advanced approaches combining AI and proteomics.
• Isomorphic Labs, a Google DeepMind spinoff, will soon begin human trials of AI-designed drugs using AlphaFold and its new IsoDDE tool, with partnerships at Eli Lilly and Novartis.
• Tencent updated its Hunyuan AI model after hiring former OpenAI scientist Yao Shunyu, and plans to double AI investment to over $5 billion this year to compete with ByteDance, Alibaba, and DeepSeek.
• Vodafone and Google Cloud launched AI security and concierge tools for small businesses, starting in Germany and Greece, as part of a $1 billion, 10-year partnership.

Intel CEO sees strong demand for AI chips and server CPUs

Intel CEO Pat Gelsinger said the company is seeing a big increase in demand for its AI accelerators and server CPUs. He noted that 14th Gen CPUs are helping to bring back server CPU demand as AI workloads grow. Gelsinger also said the PC market will decline in the second half of the year but AI-powered PCs will help offset that. Intel is also seeing interest from companies that want to build custom AI chips using Intel's foundry services. The company's gross margins have stayed around 50 percent and Intel is working to improve them further.

Intel bets on 14th Gen CPUs to win in AI market

Intel CEO Pat Gelsinger said the company is focusing on its 14th Gen CPUs to take advantage of the growing AI market. He believes Intel's manufacturing skills and customer relationships will help it meet demand for AI computing power. Gelsinger noted that CPUs are still very important for managing data and running complex AI tasks even though GPUs get more attention. Intel is also offering Gaudi accelerators and integrated CPU solutions to give customers more choices. The company is working with major cloud providers and enterprise clients to build AI solutions that fit their needs.

Harvard requires all freshmen to take AI training module

Harvard College now requires all freshmen in expository writing courses to complete a three-part training module on artificial intelligence. The module covers how large language models work, AI's impact on education, and broader questions about AI and society. It was rolled out this spring after being tested in select courses starting in fall 2024. More than 90 percent of Harvard College students take an Expos course in their first year. The goal is to give every student a shared understanding of generative AI so faculty do not have to teach the basics themselves.

Zscaler joins Project Glasswing to hide apps from AI attacks

Zscaler has joined Project Glasswing, a program led by Anthropic that gives select organizations access to the Claude Mythos Preview AI model. This model can find software vulnerabilities much faster than humans by reading code and chaining weaknesses together. Zscaler says the old approach of patching vulnerabilities after they are found no longer works against AI-powered attacks. Instead, Zscaler's Zero Trust Exchange platform makes internal applications invisible to the internet so attackers cannot find them. The platform processes over 500 billion transactions daily and uses AI to block threats before connections are made.

AI companies face hard questions about children and accountability

A CalMatters commentary argues that AI companies are causing real harm to children today, not just in hypothetical future scenarios. The author points to the US-Israeli war on Iran where Tomahawk missiles struck a school in Minab, killing 18 children. Anthropic has a contract with the Pentagon and other AI firms like OpenAI and Google DeepMind also work with the military. The piece criticizes a documentary that focused on future AI risks but ignored current harms and military contracts. It also highlights how communities in California are fighting back against harmful AI systems through strikes and local ballot measures.

OpenAI apology over Tumbler Ridge shooter raises hard questions

OpenAI CEO Sam Altman apologized for not alerting law enforcement about the Tumbler Ridge shooter's use of ChatGPT before the attack. The apology came too late and highlighted a long-standing tension between user privacy and proactive threat reporting. OpenAI has historically resisted becoming a surveillance tool and chose to offer resources to users in crisis rather than notify authorities. The Tumbler Ridge case showed the limits of that approach when a user may have shown intent to commit violence. Altman himself was recently targeted by a man who traveled to San Francisco intending to kill him, which sharpens the stakes of this policy debate.

China's auto industry races to put AI in cars as Beijing mandates

China's auto industry is moving quickly to embed artificial intelligence in cars, following Beijing's mandate in the latest five-year plan. The goal is to make the next generation of electric vehicles into self-reasoning machines that run on Chinese chips and software. China already dominates the market for electric vehicles after 25 years of development. Now the country is pushing for AI-powered cars that are not just connected to the internet but can think and make decisions on their own.

AI helps doctors predict and detect kidney disease earlier

Artificial intelligence is becoming a powerful tool in nephrology, the study of kidney diseases. Kidney diseases often develop slowly without obvious symptoms until advanced stages. AI can analyze complex data to predict disease progression and detect early signs before symptoms appear. Simple models like logistic regression and XGBoost work well for tabular medical data, while deep neural networks are used for medical image analysis. The most advanced approach combines AI with biological analysis like proteomics to find very early disease markers. Researchers emphasize that AI is a tool to support doctors, not replace them.

AI-designed drugs from DeepMind spinoff heading to human trials

Isomorphic Labs, a spinoff of Google DeepMind, will soon begin human trials of drugs designed by its AI technology. The company uses DeepMind's AlphaFold platform, which predicts protein structures and won the Nobel Prize in Chemistry in 2022. Isomorphic Labs president Max Jaderberg said the company is gearing up to go into the clinic but did not give a specific timeline. The company has developed a more powerful tool called IsoDDE for drug design. Isomorphic Labs has partnerships with Eli Lilly and Novartis for AI drug discovery.

Tencent updates Hunyuan AI model with help from former OpenAI scientist

Tencent has released a major update to its Hunyuan AI model, its first since hiring Yao Shunyu, a leading AI scientist from OpenAI. The Chinese technology company is trying to catch up with other Chinese AI developers like ByteDance, Alibaba and DeepSeek. Tencent has also backed AI startups Moonshot AI and StepFun to boost its cloud computing division. The company has restructured its research team to improve training data quality and plans to double its AI investment to more than $5 billion this year.

Vodafone and Google Cloud launch AI security and concierge for small businesses

Vodafone Business and Google Cloud have launched new AI-powered tools designed for small and medium-sized businesses. The Managed Detection and Response service uses Google's security analytics and AI to identify and stop cyber threats in real time. The Business AI Concierge uses Google's Gemini models to handle phone calls, appointment scheduling and customer inquiries automatically. The services will first launch in Germany and Greece with plans to expand to other European markets later in 2026. This is part of a $1 billion, 10-year partnership between the two companies announced in October 2024.

Sources

• Intel's AI Push: Demand Surge and Future Challenges
• Intel's AI Chip Push: Betting on 14th Gen CPUs
• Harvard College Adds Required AI Module to Expository Writing Curriculum | News
• Zscaler Is Proud to be Part of Project Glasswing: AI Can’t Breach What It Can’t Find
• Children are in the crosshairs of artificial intelligence. Who will we blame?
• OpenAI's Apology and the Line AI Companies Can No Longer Avoid
• China's auto industry races to embed AI in line with Beijing mandate
• Artificial Intelligence in Nephrology
• AI-Designed Drugs by a DeepMind Spinoff Are Headed to Human Trials
• Former OpenAI research scientist launches new AI model for Tencent
• Vodafone and Google Cloud launch AI security and concierge tools for small businesses