The unenthusiastic response to the launch of ChatGPT-5 last week has damaged faith in artificial intelligence. After all, if large language models aren’t the promised path to human-level AI, and are stuck instead on a plateau of minor tweaks, what comes next for Sam Altman and OpenAI?

This week provided an answer, when Altman’s company announced plans to invest in Merge Labs, a brain implant start-up expected to be valued at $850 million. This makes it a direct competitor to Elon Musk’s Neuralink, as well as nine other brain-computer interface (BCI) companies worldwide. Brain chips might sound like the preserve of science fiction, but they raise ethical questions around live human testing, consent, and the ultimate goals of this futuristic tech.

Musk’s Neuralink relies on invasive brain surgery via a process called PRIME: Precise Robotically Implanted Brain-Computer Interface. A robotic arm drills into the skull, placing a chip on the brain tissue. The system navigates blood vessels and inserts over 3,000 electrodes into specific brain regions. Once installed, the battery-powered chip processes neural signals, translating them into Bluetooth commands to control devices like a cursor or prosthetic limb.

The first human Neuralink implant was performed in January 2024 on quadriplegic Noland Arbaugh. The chip let him control a cursor with his thoughts, and he praised the life-changing tech, which allowed him to play chess and video games. However, the device later malfunctioned, with 85% of the electrode threads retracting from Arbaugh’s brain, disabling external control.

While seven people have received Neuralink chips and the fifth implant coincided with a $650 million funding round, no brain-computer interfaces have yet been fully FDA-approved. They have only been granted a “breakthrough device” designation, speeding up review but bypassing full scrutiny. Naturally, Neuralink’s haste has drawn criticism.

In 2022, an FDA investigation revealed “painful, deadly experiments” in Neuralink labs, where 1,500 animals — including sheep, pigs, monkeys, rats, and mice — died. One employee called the surgeries “hack jobs”, while Vox reported: “The breakneck speed at Neuralink likely caused researchers to test and kill more animals than a slower approach would.”

Marcello Ienca, a professor of AI and neuroscience ethics at TU Munich, said Neuralink “seems to sidestep established protocols”, running unregistered trials that “violate fundamental ethical guidelines”. What’s more, Musk admits that creating therapeutic aid for spinal injuries is just the first step. The real aim is “human-AI symbiosis”, which he calls “species-level important”. Beyond therapy, Musk plans to implant 20,000 people annually by 2031, tweeting: “If all goes well, we’ll have hundreds of implants in a few years, millions within a decade.” His goal is the “post-human”, a cyborg. It’s hardly an FDA-endorsed vision.

So why is OpenAI joining the brain-chip race? The Altman-backed lab, Merge, takes its name from his 2017 transhumanist concept which proposes a fusion of human brains and computers via AI. He frames it with the usual apocalyptic tech determinism: “Unless we destroy ourselves first, superhuman AI, genetic enhancement, and brain-machine interfaces will happen.”

But is Merge sincere, or just a case of smoke and mirrors? Following the failure of ChatGPT-5’, it’s easy to see why Altman might shift into brain tech to drum up another cycle of hype.

The real danger is that OpenAI’s entry into the brain-chip race could push BCI companies to cut corners in pursuit of the human-machine fusion, ending thousands of animal lives in rushed experiments, and risking human lives too. Do we really trust Musk and Altman, who are racing to win further investment, to implant millions of people? Although Altman claims his technology will be less invasive than Neuralink, the last thing we need is a race where corners are cut, and regulations and medical standards “disrupted”.

This rush to develop brain chips may receive less attention than the broader AI arms race, but it should fill us with real concern. Silicon Valley’s “move fast and break things” ethos could become far more dangerous when the things that get broken are us.