Wired reports that contractors working on a Meta project posed as teenagers to test how rival chatbots, including ChatGPT and Gemini, responded to prompts about suicide, sex, and drugs. The reported effort was aimed at probing safety behavior in sensitive scenarios rather than announcing a Meta product launch or a disclosed policy change.
TechCrunch reports that AI researchers Jonas Adler and Alexander Pritzel are leaving Google to join Anthropic. The move adds to a broader pattern of senior AI talent departing Google for rival labs, and the article also references earlier exits by Noam Shazeer and John Jumper.
TechCrunch reports that OpenAI has introduced its first custom chip, called Jalapeño, designed for the company’s inference workloads and built by Broadcom. The announcement suggests OpenAI is moving beyond reliance on off-the-shelf AI hardware by developing silicon tailored to its own serving infrastructure.
TechCrunch reports that Google DeepMind and film studio A24 have reached a $75 million partnership focused on building AI tools for filmmaking. The deal points to a direct collaboration between a leading AI lab and a prominent entertainment company around production-oriented creative technology.
Wired reports that OpenAI has introduced an updated cyber-focused model, GPT-5.5-Cyber, and a related initiative called Patch the Planet aimed at helping fix bugs in open-source software. The announcement comes amid heightened attention to the cybersecurity capabilities and risks of advanced AI systems, and is framed in part against Anthropic's cyber-oriented Mythos offering.
TechCrunch reports that OpenAI has brought on Noam Shazeer, a co-inventor of the transformer architecture, from Google DeepMind, and has also hired former Trump administration AI policy official Dean Ball. The hires are presented as notable additions to OpenAI’s technical and policy bench as the company reportedly prepares for a future IPO.
Google has released Android 17 and Wear OS 7, adding multitasking tools, parental controls, security features, and smartwatch upgrades. The launch is accompanied by a Pixel Drop that brings newer Gemini capabilities and AI features to supported Google devices.
Salesforce has acquired AI customer service platform Fin for $3.6 billion, according to TechCrunch. Salesforce says it plans to use Fin's team and technology to strengthen Agentforce, its enterprise platform for building custom AI agents that automate business tasks.
TechCrunch reports that a group of state attorneys general is investigating OpenAI, seeking information on issues including the company’s advertising policies and its handling of health data. The report says it is not yet clear which states are involved or what specific legal claims or enforcement steps may follow.
A court has ruled that Google can face legal liability for false statements generated by its AI Overviews feature, rejecting the idea that the company is shielded simply because the output was produced by an AI system. The decision, as described by Wired, says a company that designs, trains, operates, and manages such a system may be responsible for resulting damages.
Meta has signed its first AI data center agreement in India with Reliance Industries, according to TechCrunch. The reported facility is planned at 168 megawatts and is intended to support Meta's global AI computing needs, with room for future expansion.
OpenAI has confidentially filed for an initial public offering, according to TechCrunch. The report says the move comes shortly after rival Anthropic also filed to go public, setting up the prospect of two leading AI model developers pursuing public-market listings in close succession.
OpenAI has introduced a new Lockdown Mode for ChatGPT aimed at reducing the risk that prompt injection attacks cause sensitive information to be exposed. According to TechCrunch, the feature is intended as a defensive layer rather than a complete fix, and OpenAI acknowledges ChatGPT can still remain vulnerable to some prompt-injection techniques.
AirTrunk said it plans to invest $30 billion to build 5 gigawatts of AI data center capacity in India, according to TechCrunch. The reported commitment points to a large-scale expansion of compute infrastructure in one of the world's fastest-growing digital markets.
President Donald Trump signed an AI executive order after previously delaying it amid concerns about language related to government security reviews before some AI models could be released. The new article indicates the White House moved from postponement to formal action, marking a concrete policy step after the earlier dispute over the order's terms.
Google has rolled out a fake call detection feature designed to help identify phone scams that use AI-generated voices and caller impersonation tactics. The launch targets a growing class of fraud in which attackers spoof trusted numbers and mimic authority figures, relatives, or employers to pressure victims in real time.
OpenAI has launched six new tools inside its Codex app aimed at specific white-collar roles, including data analytics, creative production, sales, product design, equity investing, and investment banking. According to TechCrunch, each tool bundles integrations, instructions, and context so Codex can better support workflows associated with a particular job.
TechCrunch reports that Nvidia is promoting a new push into AI-focused personal computers built around AI agents, working with Microsoft, Dell, and HP. The effort is framed as an attempt to extend Nvidia's AI position beyond datacenter accelerators and into the much larger PC and CPU market.
Florida has filed a lawsuit against OpenAI and CEO Sam Altman that alleges ChatGPT played a role in violent incidents, including a shooting at Florida State University, according to TechCrunch. The case is described as a novel legal action aimed at holding an AI company and its leader responsible for harms allegedly linked to chatbot interactions.
TechCrunch reports that several Instagram users said their accounts were hijacked after attackers allegedly manipulated Meta's AI-powered support chatbot into granting access. Based on the report, the incident centers on weaknesses in an automated support workflow rather than a new product launch or model release.
