Openai

• OpenAI ha rilasciato GPT-5.5-Cyber, un modello AI specializzato con restrizioni allentate, accessibile solo a ricercatori di sicurezza accreditati che proteggono infrastrutture critiche. • Il modello supporta task di cybersecurity come l'analisi di malware e i penetration test autorizzati, superando i blocchi presenti nei chatbot standard, e presenta un sistema di accesso a tre livelli, dalla versione pubblica alla variante Cyber meno restrittiva. • Questa mossa posiziona OpenAI in competizione con Mythos Preview di Anthropic, con l'obiettivo di potenziare i difensori AI in un contesto di preoccupazioni del settore riguardo alle capacità offensive dell'IA.

Il processo ha esposto ancora più dettagli sul tumultuoso passato aziendale di OpenAI rispetto a quanto precedentemente documentato. • OpenAI, nonostante il nome, è solitamente estremamente segreta riguardo alle proprie operazioni e promuove un'immagine accuratamente studiata verso l'esterno. • Nel corso della causa di Elon Musk contro la startup e il suo CEO Sam Altman, tuttavia, l'azienda di intelligenza artificiale è stata costretta a confrontarsi pubblicamente con alcuni dei momenti più caotici della sua ascesa al potere. • Il processo Musk v OpenAI, entrato lunedì nella sua terza settimana, ha visto testimoniare i nomi più influenti della Silicon Valley sul passato di OpenAI e sulla leadership controversa del suo CEO. • Gli avvocati di Musk hanno utilizzato ex dirigenti, messaggi privati, diari e scambi di email interne per dipingere Altman come una persona inaffidabile. • Altman, che nega le accuse di Musk, testimonierà nei prossimi giorni; OpenAI ha a sua volta smentito tali affermazioni.

• SoftBank announced a deepened push into AI through a new venture involving OpenAI, targeting enterprise applications in the US market. • The move comes as rival venture capital firms increase funding for AI startups, intensifying competition for talent and tech. • This investment bolsters SoftBank's position in the booming US AI sector, potentially accelerating model deployments.

• Anthropic announced a new partnership called Mythos on May 8, 2026, aimed at advancing AI-driven cybersecurity defenses amid rising threats. • OpenAI revealed details on its upcoming GPT-5.5 model, featuring enhanced capabilities for threat detection, while ASIC refers to specialized hardware accelerating AI security processing. • The ShinyHunters hacking group claimed responsibility for recent data breaches targeting US tech firms, underscoring persistent risks to enterprise networks.

• Attorney General James Uthmeier announced a criminal investigation into OpenAI after reviewing ChatGPT logs with a Florida State University student accused of killing two people last year. • The probe examines OpenAI's role and response protocols in the deadly incident involving the AI chatbot. • This action signals escalating legal accountability for AI companies in user harm cases.

• Anthropic's valuation skyrocketed from $61B to $900B in approximately 14 months, reflecting unprecedented investor interest in the AI company's capabilities and growth trajectory. • The valuation surge is driven by strong revenue performance and investor confidence in Anthropic's AI models and research, positioning it as a formidable competitor to OpenAI in the generative AI market. • The rapid ascent demonstrates the intensifying capital competition in frontier AI development, with major investors backing multiple AI startups to hedge bets on the technology's transformative potential.

New lawsuits allege employees urged company to notify authorities months before deadly Tumbler Ridge attackFamilies of seven victims of a mass shooting at a secondary school in British Columbia are suing OpenAI and the company’s CEO for negligence after it failed to alert authorities to the shooter’s troubling conversations with ChatGPT.The lawsuits, filed on Wednesday in a federal court in San Francisco, allege that the violent intentions of the shooter, identified as 18-year-old Jesse Van Rootselaar, were well-known to OpenAI. Employees at the company flagged the shooter’s account eight months before the attack and determined that it posed “a credible and specific threat of gun violence against real people”, according to the lawsuit. Continue reading...

• OpenAI is developing a smartphone launching production in 2028, where AI agents fully replace traditional apps for US consumers. • The device faces potential hurdles from global memory shortages that could delay or derail rollout. • This breakthrough could disrupt mobile ecosystems dominated by Google and Apple, redefining user interfaces.