Apple Sues OpenAI in Blockbuster Trade Secret Case
The biggest story shaking Silicon Valley this week: Apple has filed a sweeping trade secret lawsuit against OpenAI in federal court in Northern California, alleging that the AI lab systematically recruited Apple employees and encouraged them to bring confidential information along.
The July 10 filing names OpenAI's hardware chief Tang Tan — a former Apple vice president — as having directed Apple employees to share company secrets during the interview process. In one striking allegation, a senior Apple systems engineer named Chang Liu allegedly failed to return an Apple-issued laptop after joining OpenAI and used it to download confidential documents about unreleased products and technologies.
"At every level, from members of its Technical Staff to its Chief Hardware Officer, and in coordination with business partners, OpenAI has been stealing Apple's trade secrets and confidential information," Apple states in its filing. The lawsuit points to more than 400 former Apple employees now working at OpenAI, many from chip design and on-device AI teams.
The case marks a dramatic reversal: Apple and OpenAI partnered in 2024 to integrate ChatGPT into the iPhone, but relations soured after OpenAI acquired Jony Ive's hardware startup IO Products for $6.4 billion and announced its own consumer hardware ambitions. Elon Musk and Sam Altman spent the weekend trading shots on X, adding fuel to what's already the tech industry's most high-profile legal battle of the year.
OpenAI responded: "We have no interest in other companies' trade secrets. We remain focused on building innovative technology that empowers people everywhere."
Sources: CNBC, TechCrunch, Fortune
OpenAI Launches GPT-5.6 and ChatGPT Work
On July 9, OpenAI released GPT-5.6 — its most capable model family yet — structured as three distinct tiers: Sol (flagship, $5/$30 per million tokens), Terra (balanced, $2.50/$15), and Luna (cost-efficient, $1/$6). All three models share a 1 million token context window and 128K max output tokens, with a February 2026 knowledge cutoff.
Sol claims state-of-the-art results across coding, knowledge work, cybersecurity, and science benchmarks, while Terra delivers comparable quality at half the cost and Luna targets high-volume production workloads. The tiered naming convention — replacing the old numbered model ladder — signals OpenAI's intention to let each tier evolve independently.
Alongside GPT-5.6, OpenAI launched ChatGPT Work, an agent that takes a goal, gathers context from connected apps, breaks the job into steps, and returns finished spreadsheets, slide decks, documents, and interactive web applications. It's built on Codex technology, which is now merging into the ChatGPT desktop app across all plans — including Free. The standalone Codex app and Atlas browser are being sunset.
Sources: OpenAI, Simon Willison, Dataconomy
Meta Launches Muse Spark 1.1 — and Its First Paid API
In a landmark shift for the company, Meta launched Muse Spark 1.1 on July 9 alongside its first-ever paid developer API. The move marks a significant strategic pivot: Meta — long the champion of open-source AI — is now charging developers to use its own model.
Muse Spark 1.1 is a multimodal reasoning model optimized for agentic tasks, with a 1 million token context window and strong capabilities in tool use, computer use, coding, and multimodal understanding. API pricing comes in at $1.25 per million input tokens and $4.25 per million output tokens — roughly 25% of what Anthropic and OpenAI charge for comparable models, as CEO Mark Zuckerberg was quick to point out.
The Meta Model API launched in US-only public preview with $20 in free credits per developer. It uses an OpenAI-compatible SDK format, making migration straightforward. On benchmarks, Meta claims Muse Spark 1.1 rivals GPT-5.5 and Opus 4.8 on agentic evaluations.
Sources: Meta AI Blog, The New Stack, Tech Times
Illinois Signs the Nation's Toughest AI Safety Law
Governor JB Pritzker signed the Artificial Intelligence Safety Measures Act (SB 315) on July 6, creating what his office calls the nation's most protective AI safety framework.
The law requires developers of large-scale AI systems to publish a safety framework detailing how they identify and assess "catastrophic risk" — defined as incidents that could cause death or serious injury to more than 50 people or more than $1 million in property damage. Key provisions include:
- Mandatory incident reporting: 72-hour deadline for safety incidents, or 24 hours if there's an imminent risk of death or serious injury
- Annual independent audits: First-in-the-nation requirement for third-party safety audits by experts without financial conflicts of interest
- Whistleblower protections: Confidential reporting channels for employees raising AI safety concerns
Combined with similar legislation in California and New York, the three states now cover roughly 40% of the U.S. AI market — creating a de facto national standard while Congress remains gridlocked on federal AI legislation. The law takes effect January 1, 2027.
Sources: Capitol News Illinois, Governing
Looking Ahead: Gemini 3.5 Pro Targets July 17
Google is reportedly preparing to launch Gemini 3.5 Pro on July 17, following a delay from its original June timeline. According to multiple reports, Google DeepMind scrapped the existing 2.5 Pro architecture for a complete rebuild.
Leaked specifications suggest the model will feature a 2 million token context window and a new "Deep Think Reasoning Layer" for complex problem-solving. However, it's worth noting that as of July 13, no official model card, pricing page, or API listing has appeared in Google's public documentation — all details remain based on third-party reporting.
If the launch holds, this week could see another major escalation in the frontier model race, with GPT-5.6, Muse Spark 1.1, and Gemini 3.5 Pro all hitting the market within days of each other.
Sources: Startup Fortune, Tech Times