Anthropic expects its suspended AI models back, and accuses Alibaba of copying its AI

On June 12, the US government ordered Anthropic to disable its two newest AI models for foreign users, citing national security. About 17 days later the models are still offline for most users, though one was restored for roughly 100 US institutions on June 26. Anthropic also told US senators that operators linked to China's Alibaba used about 25,000 fake accounts to copy its AI.

HP said it will deploy OpenAI's enterprise platform, called Frontier, across its whole company after a pilot that began in February 2026. The AI agents will handle customer support, software development, security, and internal operations across HP's business in 180 countries. HP said its security teams cut work that once took about a month down to a single day during the pilot.

LinkedIn started rolling out a tool called Premium Apply Assistant that writes and pre-fills job applications for its paying subscribers. The tool finds roles it judges a strong match, fills in application fields, and drafts cover letters. LinkedIn said recruiters see only what the applicant chooses to submit, and are not told that AI helped.

Ford said it hired, rehired, or promoted 350 experienced engineers over about three years after its automated systems failed to keep vehicle quality high. The company told reporters it had wrongly assumed AI could replace veteran staff, who now mentor junior workers and retrain its AI tools. Ford then reached the top spot among mainstream brands in J.D. Power's 2026 quality study, its best result in 16 years.

Palantir and NVIDIA announced a system that lets US government agencies run powerful AI on isolated computers that are cut off from the internet. The system uses NVIDIA's open Nemotron models with Palantir's software, and lets agencies keep control of their own data. The companies said it is aimed at government agencies and critical US industries such as energy and healthcare.