Revue de presse générée automatiquement avec
Profil généré automatiquement
Nexarion
Un individu à l'aise avec la technologie, fasciné par l'intersection de l'intelligence artificielle, de l'IA générative et de la transformation numérique, à la recherche d'informations sur les dernières tendances technologiques et leur impact sur les industries. Il suit les progrès de l'IA et de ses applications, ainsi que les tendances technologiques émergentes.
Générative ai (30%)Artificial Intelligence (30%)Technology Trends (20%)Digital Transformation (20%)
Vous souhaitez recevoir chaque jour la revue de presse de ce profil ?
OpenAI’s GPT‑5.2 launch, Disney‑OpenAI Sora partnership, AI standards and market shifts…
Vendredi 12 décembre 2025 à 10:51
Generative AI Frontiers
OpenAI unveils GPT‑5.2, a “code‑red” upgrade
OpenAI announced the GPT‑5.2 family, branding it as the most capable model for professional workloads, with claims of a 30 % drop in factual errors and breakthroughs on benchmarks such as AIME 2025. CNBC reports that CEO Sam Altman expects the rollout to quell the “code‑red” pressure by January, while Gizmodo notes the abrupt withdrawal of the prior GPT‑5.2 build amid internal alarm. MarketWatch adds that the new version is positioned as a direct counter to Google’s Gemini 3, intensifying the generative‑AI arms race.
Engadget
CNBC
Market Watch
Disney‑OpenAI Sora deal expands generative video
A three‑year licensing pact gives OpenAI’s Sora video generator access to over 200 Disney, Pixar, Marvel and Star Wars characters, effectively marrying Hollywood IP with generative‑AI creation tools. Variety highlights the $1 billion equity stake Disney is taking in OpenAI, while The Verge quotes CEO Bob Iger on responsibly extending storytelling through AI. Engadget confirms that Sora will let users generate short videos featuring icons like Mickey Mouse and Iron Man, with Disney also becoming an OpenAI API customer.
Variety
The Verge
Engadget
Agentic AI Alliance sets open‑source standards
Four AI powerhouses—Microsoft, Google, OpenAI and Anthropic—have formed the Agentic AI Alliance, a Linux‑Foundation‑backed consortium aimed at publishing open‑source standards for autonomous AI agents. Tom’s Hardware explains that the alliance will initially focus on three toolkits to address interoperability and safety challenges across agentic systems. The collaboration signals a shift toward communal governance of increasingly capable AI agents.
Tom's Hardware
Artificial Intelligence Landscape
New models raise cybersecurity alarms
TechRadar warns that the latest generation of OpenAI models carries a “high” cybersecurity risk, citing the potential for sophisticated phishing, deep‑fake generation and automated vulnerability discovery. The outlet notes that OpenAI is rolling out mitigation layers, but the escalation underscores the need for industry‑wide safeguards.
Tech Radar
State attorneys general demand safeguards against “delusional” AI output
A coalition of 42 U.S. state attorneys general has sent a letter to major AI firms—Microsoft, OpenAI, Google, Anthropic—pressuring them to implement incident‑reporting and third‑party audit mechanisms for harmful or “delusional” chatbot responses. Computer World reports that the letter, also covered by TechCrunch, seeks transparent safeguards to protect users from misleading AI behavior.
Computer World
Microsoft Copilot usage varies by device and time of day
Analysis of 37.5 million anonymized Copilot conversations reveals distinct usage patterns: desktop users favor technical assistance, while mobile users increasingly request personal advice and coaching. ZDNet emphasizes the shift toward more intimate AI interactions on mobile, whereas The Register points out that peak usage occurs in the evenings, suggesting AI is becoming a “cuddly companion” for daily life.
Zdnet
The Register
Technology Trends Shaping AI
Oracle’s AI‑heavy capex fuels lease surge
Oracle disclosed a $15 billion AI‑focused capital‑expenditure plan, prompting a steep drop in its share price. TechRadar notes investor disappointment, while CNBC reports that the company’s lease commitments have ballooned by nearly 150 % to $248 billion to support the expanding cloud and AI infrastructure.
Tech Radar
CNBC
Broadcom’s AI chip revenue jumps 74 %
The Information reveals that Broadcom’s fourth‑quarter revenue surged 28 % to $18 billion, driven by a 74 % rise in AI‑chip sales as customers like Google scale specialized processors. MarketWatch adds that this momentum helped Broadcom’s stock climb, underscoring the growing profitability of AI‑focused semiconductor portfolios.
The Information
Market Watch
Adobe credits AI for record FY2025 earnings
Adobe’s fiscal 2025 results show an 11 % year‑over‑year revenue increase to $23.77 billion, which the company attributes largely to its generative‑AI tools embedded across Creative Cloud. The Verge highlights that despite a 37 % share‑price decline, AI‑driven features such as Firefly are driving higher annual recurring revenue and cementing Adobe’s role in the global AI ecosystem.
The Verge
Digital Transformation in Motion
Rivian integrates AI for autonomous services
Rivian announced the development of proprietary AI models powering its upcoming Autonomy+ subscription, alongside an in‑house AI chip and ambitious robotaxi roadmap slated for early‑next‑year rollout. CNBC details the service’s pricing structure—$2,500 upfront or $49.99 per month—signaling a push to embed AI deeply into vehicle ownership and mobility ecosystems.
CNBC
Aller aux sources
17 sources citées