Artificial Intelligence

Bumps in the Machine: OpenAI’s GPT-5 Rollout Stumbles Into the Spotlight

In brief Developers report regressions in coding ability, raising doubts about GPT-5’s “reasoning powerhouse” claim. Rollout frustrations mount as older models vanish and access differs sharply across pricing tiers. PR stumbles and ethical concerns deepen skepticism around OpenAI’s most ambitious model yet. OpenAI’s much-hyped launch of GPT-5—touted as a groundbreaking leap in artificial intelligence—has instead…

Artificial Intelligence

Sam Altman’s OpenAI Confirms GPT-5 Is About to Drop

In brief OpenAI’s cryptic teaser shows the sign “GPT-4o” morphing into what seems to be “GPT-5.” CEO Sam Altman has been fueling the hype with memes and no-so-accidental screenshots featuring the model. Github leaks point to different GPT-5 variants touting agentic-like logic and unified architecture, with leaks benchmarks showing outstanding scores OpenAI dropped a teaser…

Artificial Intelligence

Universal Adds ‘No AI Training’ Warnings to Films as Midjourney Claims ‘Fair Use’

In brief Universal Pictures has begun adding legal warnings to film end credits stating movies “may not be used to train AI.” The warnings appear on recent releases including “How to Train Your Dragon” and “Jurassic World Rebirth.” AI firm Midjourney filed a response Wednesday to a lawsuit defending against Disney and Universal’s copyright infringement…

Artificial Intelligence

Lawmakers Call for Inquiry Into China’s DeepSeek Over National Security, Data Risks

In brief Senators cited concerns that DeepSeek’s R1 model could leak U.S. user data or aid Chinese military operations. The letter requests an investigation into potential backdoors, data flows to China, and licensing violations. Some experts warn open-source models may expose enterprises to security risks even without direct state involvement. A group of Republican lawmakers…

Artificial Intelligence

OpenAI’s ‘Jailbreak-Proof’ New Models? Hacked on Day One

In brief OpenAI released its first open-source models in years, GPT-OSS-120b/20b, hyped for jailbreaking resistance. Pseudonymous jailbreaker Pliny the Liberator cracked them in hours. Pliny’s GitHub of jailbreaks has a library of prompts to “liberate” the most important AI models. OpenAI just released its first open-weight models since 2019—GPT-OSS-120b and GPT-OSS-20b—touting them as fast, efficient,…

Artificial Intelligence

Trump Media Tests ‘Truth Search’ Using Perplexity AI

In brief Trump Media and Technology Group is working with AI giant Perplexity. The two companies have released an AI-powered search engine for President Trump’s Truth Social platform. Trump Media has a number of crypto-related ventures. Trump Media and Technology Group, which runs Trump’s social media platform, Truth Social, is working with artificial intelligence giant…