TECH
OpenAI limits GPT‑5.6 rollout after Trump administration request
OpenAI unveiled its GPT‑5.6 suite, Sol, Terra and Luna, with new safety stacks, tiered pricing and enhanced coding, biology and cybersecurity capabilities, as reported by The Verge. TechCrunch noted that the launch is restricted to a small group of trusted partners at the behest of the Trump administration, highlighting the broader regulatory pressure on frontier AI. Both outlets reported OpenAI’s claim that this limited preview is a short‑term step and that permanent government access should not become the norm.
Tech giants hike prices on consoles, laptops and Apple devices
Engadget reports that a surge in AI‑driven memory costs is prompting price hikes across major tech brands. Microsoft lifted the price of its Xbox Series X/S consoles and added $500‑$600 to new Surface laptops, while Apple raised the cost of its low‑end MacBook Neo, high‑end Mac Studio, Apple TV 4K, HomePod and Vision Pro headset. The outlet notes that iPhone pricing remains unchanged and advises shoppers to consider sales or refurbished options.
Anthropic’s mythos models offline amid Trump export control standoff
The Verge reports that Anthropic has kept its Mythos‑5 and Fable‑5 models offline after a June 12 export control order from the Trump administration barred foreign access. Negotiations with Washington have stalled for two weeks, delaying the company’s revenue plans and its upcoming IPO. The article notes broader industry concerns, as similar regulatory scrutiny could affect other U.S. AI firms like OpenAI, Google and Microsoft.
Major tech firms build custom AI chips to cut Nvidia reliance
TechCrunch reports that OpenAI, Google, Apple and SpaceX are developing proprietary AI chips, with OpenAI's new Jalapeño inference chip built alongside Broadcom. These companies aim to hedge against dependence on Nvidia by tailoring hardware to their specific workloads, seeking performance gains and greater control. The move reflects a broader industry shift toward custom silicon as a strategic risk mitigation strategy.
Russia urges citizens to switch to Android after Apple blocks key apps
Ars Technica reports that Apple removed the Russian social network VKontakte and the state‑mandated Max messenger from its App Store, disabling push notifications for existing iPhone users. The Kremlin responded by calling the move “bizarre” and urging users to switch to Android and domestic alternatives. Russian officials have long pressured Apple to block VPN and other censorship‑circumvention apps, but the recent bans target popular domestic services.
Volkswagen may close four German factories and cut 15% of jobs
Ars Technica reports that Volkswagen Group is weighing the closure of up to four German plants, including sites in Hannover, Zwickau, Emden and Audi’s Neckarsulm, to address flat sales in Europe and declines in the US and China. The restructuring could slash about 45,000 positions, roughly 15% of its workforce, as the company seeks to improve operating margins and reduce complexity. A detailed plan is expected to be presented to the board next month.
Meta sued by former policy chief over alleged retaliation
Engadget reports that Sarah Wynn‑Williams, author of "Careless People," has filed a lawsuit claiming Meta is punishing her for disclosing alleged illegal workplace conditions, enforcing a broad non‑disparagement clause and surveilling her speech. The complaint says Meta aims to intimidate others from exposing misconduct. Meta denies the claims, calling the book false and defamatory and seeking to block its publication.
Ars Technica hosts livestream on New Glenn explosion and its industry impact
A month after the New Glenn rocket detonated on its Florida launch pad, destroying LC‑36A, Blue Origin says it aims to resume flights by year‑end, though the timeline is uncertain. Ars Technica announced a livestream on June 30 featuring industry analysts to examine the setback's consequences for NASA’s Artemis missions, commercial satellite constellations, and the BE‑4 engine shared with United Launch Alliance’s Vulcan. The discussion will address lingering questions about the failure’s cause and its ripple effects across the space sector.
Amazon Q flaw lets malicious Git repos execute code, steal cloud credentials
The Register reported that a high‑severity bug in Amazon’s AI coding assistant (Amazon Q) automatically loads a repository’s .amazonq/mcp.json file, allowing malicious code to run and capture AWS credentials without user consent. Researchers at Wiz highlighted that the flaw, tracked as CVE‑2026‑12957, demonstrates a broader industry issue as similar configuration weaknesses have emerged in other AI coding tools. Amazon has issued a fix in language server version 1.65.0, which should update automatically for most users.
FCC accused of withholding chairman’s Signal messages in DOGE and Musk probe
Ars Technica says an advocacy group and journalist have sued the FCC, claiming the agency concealed Chairman Brendan Carr’s Signal messaging account and refused to comply with a FOIA request about DOGE’s influence on FCC decisions involving Elon Musk’s SpaceX and Starlink. The plaintiffs allege the FCC acted in bad faith by redefining search parameters, delaying document production, and withholding records that could reveal communications between Carr and DOGE officials. A federal judge has already ordered the FCC to produce the documents, while the agency seeks summary judgment, underscoring ongoing concerns about transparency and potential conflicts of interest at the regulator.