Media

AI news flash: Conno Christou, a 35-year-old entrepreneur, spent years cross-tracking sleep with Whoop and Oura, undergoing nearly 100 biomarker blood tests annually, and rigorously following health protocols from longevity researchers like Peter Attia and Rhonda Patrick for four years, meticulously managing supplements, circadian rhythms and protein intake. His 2025 checkup showed all green lights — the best shape in years. Then a post-workout arm swelling led to a blood clot discovery, and a pre-surgery scan revealed an 11×11×8 cm tumor behind his sternum. Biopsy confirmed aggressive non-Hodgkin lymphoma, a random genetic mutation with no link to lifestyle, that had grown in just three months.

AI news flash: Apple’s VP Paul Meade, who heads the Vision Pro headset, is leaving to join OpenAI’s hardware team, according to Bloomberg reporter Mark Gurman. Meade led Vision Pro development and reportedly headed Apple’s AI smart glasses project planned for next year. Vision Pro’s high pricing led to sales falling short of expectations, and Apple hopes more affordable smart glasses can compete with Meta’s wearables.

AI News Flash: The US government is implementing case-by-case review controls on frontier AI model releases, with the impact spreading from Anthropic to OpenAI. Following the government’s halt of Anthropic’s Fable and Mythos models, OpenAI’s latest GPT 5.6 faces the same fate — according to The Information, the model only has access to limited preview, requiring individual government approval for client access before formal release is considered.

AI News Flash: Nvidia has dominated the AI chip market for years, but this highly concentrated reliance is showing signs of loosening. OpenAI recently unveiled its in-house inference chip project, codenamed ‘Jalapeño’, developed in partnership with semiconductor giant Broadcom, specifically optimized for inference-stage computational demands. This move officially places OpenAI in the camp of homegrown chip developers alongside Google, Apple, and SpaceX.

AI News Flash: OpenAI continues to double down on India market, announcing appointment of former Uber India & South Asia President Prabhjeet Singh as India’s first Managing Director. Singh will officially assume the role this September, reporting directly to OpenAI APAC Managing Director Kiran Mani, responsible for overall business covering consumer growth, enterprise adoption, partner relationships, regulatory communication, and overall operations.

AI News Flash: OpenAI recently implemented access restrictions on GPT-5 and GPT-6 rollout following a US government request. OpenAI publicly stated strong opposition to making such government-led access review mechanisms the long-term default. OpenAI warns that these restrictions will prevent the most capable AI tools from reaching those who truly need them, affecting general users, app developers, enterprise customers, cybersecurity defenders, and global partners.

AI News Flash: Trump’s administration shows signs of softening stance on Anthropic’s cybersecurity models Mythos 5 and Fable 5, which were banned two weeks ago. Commerce Secretary Howard Lutnick sent a letter to Anthropic’s Chief Operating Officer Tom Brown on Friday, confirming that necessary safeguards are in place, allowing over 100 US government agencies and companies to access Mythos 5…

AI news flash: Based on anonymized transaction data from Indagari tracking over 28 million US consumers, paid Anthropic Claude users are growing steadily. The data covers weekly transactions from 2025 through May 10, 2026, including subscriptions and API token purchases, showing Claude’s paid user count and revenue has grown month-over-month since January 2026, with a cumulative increase of approximately 75%.

AI news flash: Patronus AI is a San Francisco startup founded in 2023 by former Meta AI researchers Anand Kannappan and Rebecca Qian, specializing in creating simulated digital environments for AI model providers and developers to assess AI agent reliability in complex real-world scenarios. Its core technology called Digital World Model can replicate websites and enterprise internal systems, allowing AI agents to undergo stress testing in these simulated environments. The training approach uses reinforcement learning, rewarding successful task completion and penalizing errors to iteratively improve agent performance.

AI News Flash: OpenAI’s latest model GPT-5.6 will be released in a completely different way than before. According to The Information, OpenAI CEO Sam Altman told employees this week in an internal meeting that the Trump administration is requiring the company to “review each customer’s access eligibility on a case-by-case basis” during the preview period, with no public release and limited to a few specific partners. If this limited release goes smoothly, OpenAI hopes to do a broader public release “in a few weeks”.