AI Daily Roundup: OpenAI Slashes Prices 80%, Judge Blocks Anthropic Ban, Okta Buys Permiso, Google's AI Bug Bounty, and FCC Bans Chinese Robots
A federal judge deals a blow to Trump's Anthropic ban, OpenAI goes on the offensive with massive price cuts, Okta makes its biggest AI security acquisition, Google proves AI can find bugs faster than humans, and the FCC locks Chinese robots out of the US market.
1. OpenAI Cuts GPT-5.6 Luna Prices by 80% in Aggressive China-Style Pricing Move
In its most dramatic pricing move yet, OpenAI is slashing the price of its GPT-5.6 Luna model by 80 percent and the Terra model by 20 percent, effective July 30. The company credits efficiency gains from its top-tier Sol model for enabling the cuts — essentially using its most advanced system to optimize the infrastructure that runs its cheaper models.
The move signals OpenAI is fully embracing a volume-over-margin strategy, directly targeting the price pressure coming from Chinese open-source models that have been undercutting Western AI providers for months. For developers and enterprises building on OpenAI's API, the cuts make the platform dramatically more competitive against open-weight alternatives. The question now is whether Anthropic and Google will follow suit.
2. Judge Rules Trump Administration Still Lacks Evidence for Anthropic 'Supply-Chain Risk' Label
A federal judge has ruled that the Trump administration has not presented sufficient evidence to justify its designation of Anthropic as a supply-chain risk — a label that effectively banned the company's AI technology from government use and cast a shadow over its commercial prospects. The ruling is a significant legal win for Anthropic and raises questions about the evidentiary standards behind government AI restrictions.
The case has been closely watched as a bellwether for how far the executive branch can go in restricting AI companies based on national security claims. With the judge casting doubt on the government's case, other AI companies facing similar scrutiny may find new leverage in pushing back against politically motivated designations.
3. Okta Acquires AI Security Startup Permiso for Approximately $200M
Identity giant Okta has acquired Permiso, an AI security startup specializing in identity threat detection, for approximately $200 million according to sources familiar with the deal. The acquisition gives Okta critical capabilities for securing AI agents and other non-human identities across cloud environments — a rapidly growing need as enterprises deploy autonomous AI systems.
The deal reflects a broader industry trend: as AI agents become more autonomous and widespread, the attack surface for identity-based threats expands dramatically. Traditional identity and access management tools weren't designed for non-human actors that can create accounts, request permissions, and move laterally through systems. Permiso's technology addresses exactly this gap, positioning Okta at the center of the emerging AI identity security market.
4. Google Fixed More Chrome Bugs in June Than the Past Two Years Combined, Thanks to AI
Google has revealed that it patched more Chrome security vulnerabilities in June 2026 alone than it did over the previous two years combined — and it credits AI tools for the dramatic acceleration. Using large language models and AI-powered code analysis, Google's security team has been able to identify and fix an exponential number of bugs across its browser codebase.
The disclosure validates a pattern that security researchers have been warning about: AI is a force multiplier for both defenders and attackers. While companies like Microsoft and Google are using AI to find and patch bugs faster than ever, the same capabilities are available to threat actors. The net effect is an arms race in software security that is accelerating the pace of vulnerability discovery on both sides.
5. FCC Bans New Chinese Robots and Power Inverters to Protect US AI Infrastructure
The FCC has issued a sweeping ban on imports of new Chinese humanoid robots, robot dogs, and power inverters, citing supply chain vulnerabilities and cybersecurity risks to the US AI buildout. The rule's broad definition, however, also sweeps in consumer devices like robot vacuums, robotic lawn mowers, and delivery bots — a detail that has caught manufacturers off guard.
The commission's report highlights specific security incidents, including vulnerabilities in foreign robots that allowed remote access to thousands of home devices with live camera feeds and floor plans, and pre-installed backdoors found in Chinese robot dogs. The ban only applies to models not yet released; existing devices, self-driving cars, drones, surgical robots, and fixed robotic arms are exempt. However, since virtually all robot vacuums come from China — including market leaders Roborock, Ecovacs, and Dreame — the impact on the consumer market could be significant. The Department of War will handle exemption decisions going forward.