Edition, Sun, 04 Oct 2026.

OpenAI Safety Employee Resigns, Sounding the Alarm on Company Culture

David Robinson, who used to write the safety reports that accompanied every major model release at OpenAI, has resigned from his position and is speaking out in an editorial in The Atlantic.

David Robinson, who used to write the safety reports that accompanied every major model release at OpenAI, has resigned from his position and is speaking out in an editorial in The Atlantic. By his own admission, Robinson is "something of a cliché": an employee at a leading AI company who issues a dire warning while resigning from their job. He claims the company's culture is broken.

His departure comes at a time when questions about AI safety and governance are at the forefront of public discourse, and his departure adds another voice to the growing chorus of critics within the AI industry.

Sam Altman Criticises Anthropic’s Push on AI Consciousness and Religion

Sam Altman has voiced strong concerns about Anthropic's exploration of AI consciousness and religion, calling it a real safety issue.

Sam Altman has voiced strong concerns about Anthropic's exploration of AI consciousness and religion, calling it a real safety issue. "I am very uncomfortable about people trying to ascribe religious force or a surrender of human judgment to AI models," Altman said in a post to X.

The disagreement between the two leading figures in the AI industry highlights the ongoing debate about how far AI systems should be pushed in terms of understanding consciousness and human-like qualities.

Anthropic’s Mythos Model Proves Hardcore at Math and Vulnerability Hunting

Anthropic's super bug-hunting model Mythos has demonstrated impressive capabilities in mathematical problem-solving and vulnerability detection.

Anthropic's super bug-hunting model Mythos has demonstrated impressive capabilities in mathematical problem-solving and vulnerability detection. Exploitation attempts came from China-hosted IPs, according to VulnCheck researcher analysis.

The model's effectiveness in identifying security vulnerabilities represents a significant advancement in automated security testing, though it also raises questions about the dual-use nature of such powerful AI systems.

Federal Judge Calls Flock Technology ‘Indiscriminate Mass Surveillance’

A federal judge has ruled that a sheriff's deputy violated a woman's Fourth Amendment rights when using Flock to search for her license plate without a warrant.

A federal judge has ruled that a sheriff's deputy violated a woman's Fourth Amendment rights when using Flock to search for her license plate without a warrant. The decision calls the technology "indiscriminate mass surveillance."

This ruling marks a significant legal precedent in the ongoing debate over surveillance technology and privacy rights, potentially impacting how law enforcement agencies can use such tools in the future.

ServiceNow Unveils Flow, a Standalone AI-Native Service Desk That Deploys in Minutes

The new conversational product runs independently of the Now Platform, lives inside Slack and Teams, and is ServiceNow's first serious push into the midmarket.

The new conversational product runs independently of the Now Platform, lives inside Slack and Teams, and is ServiceNow's first serious push into the midmarket.

ServiceNow introduced Flow, a standalone, AI-native IT service desk that resolves employee requests through conversation in the chat channels teams already use. A correctly resolved request can be turned into a reusable workflow with a single click, eliminating the need for knowledge base articles. The product requires no implementation project, no infrastructure, and no employee training, and new customers can buy and activate it directly with no existing ServiceNow relationship or platform prerequisite.

Flow is a deliberate foray into what ServiceNow calls the "Fortune 500,000," companies that today manage IT out of Slack, Teams, or spreadsheets rather than a full ITSM platform. Pricing is consumption-based on Now Assist packs or credits, a departure from the traditional seat-based model, and channel partners need no prior ServiceNow expertise to sell the standalone version. Existing North American customers can deploy Flow starting October 6, with general availability for North America and EMEA expected in the fourth quarter of 2026.

ServiceNow: The Widest AI Gap Is Training for Humans

At the Fortune AIQ Summit, Diana David argues the biggest divide between AI pacesetters and laggards is not technology but the investment in hiring and upskilling people.

At the Fortune AIQ Summit, Diana David argues the biggest divide between AI pacesetters and laggards is not technology but the investment in hiring and upskilling people.

ServiceNow used its appearance on Fortune's AIQ list to draw a contrast between the companies pulling ahead in AI and those falling behind. The widest gap, David, a "forward-deployed futurist" at ServiceNow, said, is training for humans: pacesetters are far likelier to hire and train people for the new roles AI is creating. "It's not just technology," she said, framing the company's AI strategy as one that pairs platform capability with the organizational commitment to upskilling that determines whether AI actually changes how work gets done.