AI Patents, Driverless Uber, and Chip Showdown

Welcome to AI News in 10, your top AI and tech news podcast in about 10 minutes. AI tech is amazing and is changing the world fast, for example this entire podcast is curated and generated by AI using my and my kids cloned voices...

It’s Thursday, November 27, 2025. Here’s what’s new in AI and tech today... The U.S. Patent and Trademark Office issued fresh guidance on how to handle inventions created with AI assistance... Uber and China’s WeRide flipped the switch on fully driverless robotaxis in Abu Dhabi — the first outside the U.S. on Uber’s platform... HP is planning thousands of job cuts as it retools around AI... OpenAI disclosed a security incident at its analytics vendor Mixpanel but says no ChatGPT users were affected... and the chip world is buzzing as Intel denies TSMC’s trade-secret allegations while Taiwan steps up its probe. Let’s get into it.

[BEGINNING_SPONSORS]

Story one — patents and AI.

On Wednesday, November 26, the U.S. Patent and Trademark Office set out new rules for AI-assisted inventions. The headline is simple: AI is a tool, not an inventor.

Under the guidance, only a human who conceived the invention can be named as the inventor, even if generative AI helped along the way. The agency rejected a special test for AI-assisted work — the same inventorship standards apply whether you used a lab notebook, a microscope... or a chatbot. This also reaffirms prior court rulings that an AI system cannot be listed as an inventor under U.S. law.

Why this matters... It tells companies and researchers that using AI won’t automatically disqualify patents — but you’ll need clear documentation showing the human contribution. Expect in-house counsel to tighten disclosure forms and lab protocols to track who did what, when, and how, especially in fields like drug discovery and materials where AI tools are now routine.

A quick implication check. Startups leaning on autonomous agents for ideation should be careful about over-automating the conception step. If your product pipeline is prompt-driven synthesis with minimal human reasoning, you may struggle to prove inventorship. Conversely, teams that treat AI like any other instrument — and can show human insight guided the claims — should feel more confident about filing. Aligning with existing standards also reduces the risk of a patchwork, AI-only inventorship doctrine that could have increased litigation.

Story two — Uber goes driverless... in the Middle East.

Uber and WeRide have started fully driverless robotaxi rides in Abu Dhabi, with public operations kicking off on Yas Island. Riders requesting UberX or Uber Comfort can be matched with a WeRide robotaxi, and there’s even a new “Autonomous” option to boost your chances of getting a vehicle with no one behind the wheel. It’s the first fully driverless service on Uber’s platform outside the U.S., enabled by a city-level permit from Abu Dhabi’s Integrated Transport Centre.

The companies say they’ll expand coverage beyond the initial zone by year’s end, and WeRide currently operates more than 100 robotaxis across the region. This is a notable milestone for Uber’s strategy to be the neutral marketplace that aggregates autonomous fleets — not just a human-driver app.

What’s different here compared with other robotaxi pilots? Two things. First, true driverless service — no safety operator — which is still rare globally. Second, the marketplace model: Uber handles rider matching and operations logistics with local fleet partner Tawasul, while WeRide focuses on the autonomy stack and vehicle tech. If unit economics approach breakeven, as the partners suggest, that would signal that constrained-area, tourism-heavy districts — like Yas Island — are a pragmatic path to scale. The caveat, of course, is expansion: can they move beyond geofenced areas and into denser city cores on a predictable timeline? We’ll be watching for service maps to widen and for regulators to issue broader permits.

Story three — HP trims to fund the AI pivot.

HP says it will cut between 4,000 and 6,000 jobs by the end of fiscal 2028 — up to about 10% of its workforce — as part of a plan to streamline operations and embed AI more deeply across product development, support, and internal processes. The company expects around a billion dollars in annual run-rate savings by 2028, though it will take roughly $650 million in restructuring charges to get there. The decision came alongside mixed guidance and follows an earlier, smaller reduction in February. It’s another data point in a broader 2025 pattern — companies touting AI investment while rebalancing headcount to pay for it.

Here’s the tension. The PC market is positioning for an “AI PC” upgrade cycle, but upstream components — especially memory — are in tight supply due to data-center demand. That squeezes margins for device makers even as they push new AI features. If memory prices keep rising into 2026, PC and workstation pricing may creep up, too. HP’s move suggests incumbents will restructure to preserve cash for AI research and development and go-to-market, while leaning on partners for supply commitments. For employees, the cuts will be felt across product, ops, and support — roles that companies increasingly augment with AI copilots and agents.

[MIDPOINT_SPONSORS]

Story four — security watch: OpenAI and Mixpanel.

OpenAI disclosed a security incident at Mixpanel, the analytics vendor it used for the frontend of its API platform. According to OpenAI, an attacker exported a dataset from Mixpanel that included limited, non-sensitive analytics information for some API users — names, email addresses, approximate location, operating system and browser metadata, and certain identifiers. Crucially, there’s no evidence of a breach of OpenAI systems, and no chat content, API prompts or outputs, passwords, API keys, or payment details were exposed.

OpenAI has removed Mixpanel from production, is notifying affected users, and is warning about possible phishing. For most listeners, the headline is simple: if you’re an API customer, watch for targeted spam; if you’re a ChatGPT consumer user, you weren’t affected.

Mixpanel says it detected a smishing attack on November 8 and performed resets, forensic review, and law-enforcement notifications. The company published a summary of the steps it took post-incident. Independent reporting today reiterates that ChatGPT consumer users weren’t impacted, but that some developer-profile data could be used for phishing. The broader lesson here is third-party risk — even if first-party systems are locked down, telemetry and analytics integrations can become side doors. Expect more vendors — and their customers — to revisit which analytics run where, how data is minimized, and whether privacy-preserving measurement can replace broad event capture.

Story five — a semiconductor flashpoint: Intel and TSMC.

Today, Intel publicly denied allegations from TSMC that a senior executive who recently rejoined Intel leaked trade secrets after a long career at TSMC. TSMC has filed suit in Taiwan’s Intellectual Property and Commercial Court, and Taiwan’s economic authorities say they’re looking into potential violations of laws protecting key technologies. Later in the day, Taiwan prosecutors said investigators searched two homes tied to the executive and seized computers and storage devices as part of an ongoing probe. Intel says it enforces strict policies against misuse of third-party IP and has no reason to believe there’s merit to the allegations.

Why this matters. As chips get more advanced — and as national policy wraps itself around semiconductor leadership — the line between legal talent mobility and illegal trade-secret transfer is razor thin. Taiwan is particularly sensitive about protecting know-how in nodes like 3 nanometers and 2 nanometers. For Intel, which is racing to catch up in foundry manufacturing and courting partners and customers, the optics are delicate — for TSMC, signaling vigilance protects its crown jewels. Regardless of the case’s outcome, expect tighter agreements adjacent to non-competes, more frequent audits of employee devices, and closer government scrutiny when top engineers switch teams.

Quick hits to close us out...

In Europe, lawmakers passed a non-binding resolution urging an EU-wide minimum age — 16 — for access to social platforms, video sites, and AI chatbots, with parental consent between 13 and 16. It’s not law yet, but it signals a push for harmonized youth protections online, and could influence how app makers design teen experiences and parental controls.

And circling back to the U.S., the White House’s “Genesis Mission” executive order, signed Monday, aims to build a national AI experimentation platform across Department of Energy labs and supercomputers to accelerate scientific discovery — from protein modeling to fusion simulations. Watch for agency RFPs and public-private partnerships to spin up around this.

That’s the rundown: the USPTO tells innovators to keep humans on the inventorship line... Uber’s driverless rides get real in Abu Dhabi... HP restructures to fund the AI future... OpenAI’s vendor incident is a reminder to harden your third-party stack... and the Intel—TSMC dispute shows just how high the stakes are in the chip race. We’ll be back tomorrow with more. Until then, stay curious — and stay safe out there.

Thanks for listening and a quick disclaimer, this podcast was generated and curated by AI using my and my kids' cloned voices, if you want to know how I do it or want to do something similar, reach out to me at emad at ai news in 10 dot com that's ai news in one zero dot com. See you all tomorrow.