OpenAI has announced that its latest model, GPT-5.2, has set a new benchmark for professional knowledge work, outperforming industry professionals across 44 occupations in a benchmark test. The company claims that GPT-5.2 can complete tasks over 11 times faster and at less than 1% of the cost of expert professionals, making it a significant tool for professionals. This release follows a “code red” declaration from CEO Sam Altman, who emphasized the urgency of enhancing ChatGPT’s capabilities in response to increasing competition, particularly from Google’s Gemini 3. In internal tests, GPT-5.2 showed a 9.3% improvement in spreadsheet modeling tasks for junior investment banking analysts compared to its predecessor, GPT-5.1. The new model is set to roll out for paid users starting today, with the goal of improving productivity and efficiency in various professional settings.
And then The Walt Disney Company announced a significant investment of $1 billion in OpenAI, which will allow the artificial intelligence company’s Sora video generation tool to incorporate over 200 characters from Disney, Marvel, Pixar, and Star Wars. This three-year licensing agreement aims to merge Disney’s iconic storytelling with cutting-edge AI technology, providing users the ability to create personalized social videos. Bob Iger, Disney’s CEO, emphasized that the company’s collaboration with OpenAI is designed to respect and enhance human creativity, rather than threaten it. This partnership provides a framework for integrating artificial intelligence into Disney’s offerings, allowing consumers to engage with beloved characters in innovative ways while ensuring creators are honored through licensing fees. During a CNBC interview, Iger noted that the deal includes approximately 200 Disney characters and props, enabling users to immerse themselves in scenes without using the characters’ voices. Wired has a really good dive into the copyright implications if you want to explore that.
Disney then filed a cease-and-desist letter against Google, alleging “massive” copyright infringement related to the unauthorized commercial distribution of images and videos featuring characters from its popular franchises. According to the letter, Google operates as a “virtual vending machine” that reproduces and distributes copyrighted content from Disney’s extensive library, including characters from “Frozen,” “The Lion King,” and “Moana.” Google has not confirmed the allegations but stated that it values its relationship with Disney and is willing to engage in discussions.
Oh, and OpenAI says this is all very dangerous, issuing a warning regarding its forthcoming artificial intelligence models, stating they could pose a “high” cybersecurity risk. The company highlights concerns that these advanced models may develop zero-day exploits or assist in complex intrusion operations targeting real-world systems. To mitigate these risks, OpenAI is investing in strengthening its models for defensive cybersecurity tasks and plans to introduce a program offering tiered access to enhanced capabilities for users involved in cyber defense. Additionally, OpenAI will establish the Frontier Risk Council, which aims to collaborate with experienced cybersecurity professionals to address these challenges and expand into other capability domains in the future.
Not to be outdone, Google unveiled its latest AI research agent, the Gemini Deep Research. The new Gemini agent, built on the advanced Gemini 3 Pro model, enhances developers’ capabilities by allowing them to embed its research functions into their applications via the newly introduced Interactions API. This AI tool excels at synthesizing large amounts of information for diverse tasks, from due diligence to drug toxicity research. Google claims that the agent minimizes inaccuracies, known as AI hallucinations, which are critical in complex reasoning tasks. The company has also launched a new benchmark, DeepSearchQA, to evaluate performance on intricate information-seeking tasks.
Why do we care?
Let’s strip the drama out of this.
“Code Red” isn’t a thing — focus is. And this is what focus looks like when the market tightens and competition gets real.
The benchmarks sound impressive until you remember they’re measuring slices of work, not jobs. Faster spreadsheets don’t equal replaced analysts. That gap still matters.
The Disney deal? That’s the real signal. Licensing was always the endgame. Big IP doesn’t get disrupted — it gets paid. And notice how Disney licenses on one side and sues on the other. That’s not hypocrisy; that’s leverage.
Here’s the uncomfortable part: smaller companies don’t have Disney’s enforcement muscle. Copyright in the AI era is going to be uneven, and that risk lands on customers unless someone helps them navigate it. The real risk isn’t that customers make bad AI decisions — it’s that they don’t realize they’re making decisions at all. Features ship enabled. Models get bundled. Workflows change quietly. And when something breaks, the customer still owns the outcome.
And the security warning? That’s future-tense fear. If it’s dangerous later, tell me when and how — not just that it sounds scary.
So what should MSPs do? Stop selling hype. Start selling clarity. Models are interchangeable. Governance, integration, and risk management aren’t. That’s where the real value is now.