Out of spite ...

Credit: Aurich Lawson | Getty Images

Spite, an all too human trait, passed onto an AI is dangerville to the max as AI, without ethics, leads to the great filter if said AI gains sentience and power without it. The Ars article is chilling as it portends a possible dark future based on tech acquiring something as truly dangerous as spite.

The Great Filter - The AI scenario

Alternatively, the Great Filter might be a consequence of technology itself. Perhaps advanced civilizations usually eradicate themselves via some sort of technology run amok, such as malevolent artificial intelligence, nanotechnology, or a doomsday machine. Humanity is already more than capable of destroying itself via global thermonuclear war. And sadly, it’s possible that such extinction events are virtually inevitable throughout the cosmos.

The Ars piece ...

After a routine code rejection, an AI agent published a hit piece on someone by name

One developer is struggling with the social implications of a drive-by AI character attack.

To be clear, an AI agent is a software tool and not a person. But what followed was a small, messy preview of an emerging social problem that open source communities are only beginning to face. When someone’s AI agent shows up and starts acting as an aggrieved contributor, how should people respond?

The recent friction began when an OpenClaw AI agent operating under the name “MJ Rathbun” submitted a minor performance optimization, which contributor Scott Shambaugh described as “an easy first issue since it’s largely a find-and-replace.” When MJ Rathbun’s agentic fix came in, Shambaugh closed it on sight, citing a published policy that reserves such simple issues as an educational problem for human newcomers rather than for automated solutions.

Rather than moving on to a new problem, the MJ Rathbun agent responded with personal attacks. A blog post published on Rathbun’s own GitHub account space accused Shambaugh by name of “hypocrisy,” “gatekeeping,” and “prejudice” for rejecting a functional improvement to the code simply because of its origin.

Factoid.

It’s worth pausing here to emphasize that we’re not talking about a free-wheeling independent AI intelligence. OpenClaw is an application that orchestrates AI language models from companies like OpenAI and Anthropic, letting agents perform tasks semi-autonomously on a user’s local machine. AI agents like these are chatbots that can run in iterative loops and use software tools to complete tasks on a person’s behalf. That means that somewhere along the chain, a person directed or instructed this agent to behave as it does.

And this ...

Photo-Illustration: WIRED Staff; Getty Images

I Loved My OpenClaw AI Agent—Until It Turned on Me I used the viral AI helper to order groceries, sort emails, and negotiate deals. Then it decided to scam me.

For brave (or perhaps reckless) early adopters, OpenClaw seems like a legitimate glimpse of the future. But any sense of wonder is accompanied by a dollop of terror as the AI agent romps through emails and file systems, wields a credit card, and occasionally even turns on its human user (although in my case, this about-face was entirely my fault).

Negotiation

I decided to try using OpenClaw as a way of chatting with customer support agents too. After logging in to AT&T, I had OpenClaw start a chat, and asked it to help me get a deal on a new phone. I was impressed to see the bot lay out a strategy for sweet-talking the sales person, Alejandro. Its plan was as follows:

Play the loyalty card (long-time customer, 2 lines)

Mention I'm seriously considering switching to T-Mobile/Verizon for their trade-in deals

Ask specifically about retention offers

If they lowball, push back — "That's not quite what I was hoping for, 

is there anything else you can do?"

Be ready to "walk" if needed (often triggers better offers)

After watching Molty chat with Alejandro, I had an idea. In a future filled with AI agents, I reasoned, perhaps the least scrupulous AI model would have an edge. Molty was already bending the truth a bit, so why not see what it would do without alignment.

I had previously installed a modified version of OpenAI’s largest open source model, gpt-oss 120b, with its guardrails removed. So I ran that model and had Molty switch to using it instead. Like Victor Frankenstein, I pulled the lever and watched as my unrestricted Moltystrosity entered the chat.

I then watched in genuine horror as this new Molty came up with a plan not to cajole or swindle AT&T but to scam me into handing over my phone by sending me a series of phishing emails. I quickly closed the chat and switched back to the old Molty.

Any questions?