Microsoft, OpenAI Execs Acknowledge AI Theft of Labor

Key Takeaways

• Court filings reveal Microsoft and OpenAI executives privately called AI training the largest theft of labor in history
• Internal Microsoft data showed Copilot reduced New York Times referrals by up to 93 percent compared to Bing
• OpenAI head Nick Turley warned AI products pose an existential threat to news publishers
• HP uncovered a fake AI crypto trading tool distributing Needle Stealer malware targeting MetaMask, Coinbase Wallet, and Phantom
• Meta launched Meta One subscription plans ranging from $7.99 to $499 per month to fund $130-145 billion in annual capex
• Gerstner says AI labs need combined revenue to grow from $100 billion to $180 billion by year-end
• Anthropic opened a physical biology lab in the Bay Area for AI-driven drug research
• Mastercard developed Verifiable Intent to record who authorized AI agents to make purchases on behalf of consumers
• K&L Gates adopted AltaClaro's DepoSim for AI-powered deposition training across its litigation practice
• Only 7% of U.S. and U.K. consumers said they would let an AI assistant make purchases without approval

OpenAI and Microsoft executives' quotes undermine AI fair-use defense

Court filings made public on September 17 show OpenAI and Microsoft executives privately described their AI products as substitutes for journalism. These quotes weaken the tech companies' fair-use defense in a copyright lawsuit filed by the New York Times and other news outlets. The lawsuit accuses the companies of using millions of articles without permission to train ChatGPT. Quotes from executives including Greg Brockman, Satya Nadella, and Nick Turley were revealed in the filings. The case will likely hinge on whether AI training counts as fair use of copyrighted material.

Microsoft executive called AI training the largest theft of labor in history

Unredacted court filings revealed that Microsoft and OpenAI executives privately called AI training theft while publicly defending it as fair use. Microsoft's director of applied science Brent Hecht called it the largest theft of labor in human history. Internal Microsoft data showed Copilot produced up to 93 percent fewer referrals to The New York Times compared to Bing search. OpenAI head Nick Turley warned that publishers face an existential threat from substitutive AI products. The filings were made public on September 17, 2026, in The New York Times copyright case.

Tech staffers knew AI tools posed existential threat to news publishers

Court documents reveal that high-level staffers at Microsoft and OpenAI knew AI training could devastate news publishers but continued anyway. OpenAI head Nick Turley warned that AI products pose an existential threat to publishers and would become more substitutive over time. Microsoft's Brent Hecht described a doom loop where AI hurts the performance of models and the web. The filings came from a copyright case brought by The New York Times and other news outlets. The documents also described OpenAI discussing ways to skirt publishers' paywalls.

Microsoft memo called AI scraping largest theft while OpenAI threatened publishers

Legal briefs in The New York Times lawsuit against OpenAI and Microsoft revealed damning internal documents. A January 2023 memo by Microsoft's Applied Science director Brent Hecht called AI scraping the largest theft of labor in human history. The same memo noted Copilot reduced click-through rates to The New York Times by up to 93 percent compared to Bing. OpenAI head Nick Turley described ChatGPT as a modern newsstand and a threat to publishers. The revelations came from newly unredacted filings made public on September 17.

Microsoft memo labels AI training largest theft of labour in history

A Microsoft internal memo from science director Brent Hecht called using copyrighted material to develop ChatGPT astonishing and the largest theft of labour in history. The document was revealed in The New York Times's lawsuit against OpenAI and Microsoft. Hecht's memo also described a doom loop where AI products threaten the economic foundations of content suppliers. The filings showed internal Microsoft data predicting harmful outcomes before the products were shipped. Microsoft spokesperson said Hecht's views do not represent the company's position.

Fake AI crypto trading tool steals wallet passwords

Cybercriminals created a fake AI crypto trading agent that installs malware called Needle Stealer. The trick replaces real wallet extensions like MetaMask, Coinbase Wallet, and Phantom with malicious copies. Victims who enter their passwords send them directly to attackers. The campaign was discovered by HP in a report dated September 17, with activity noted in April and June 2026.

Fake AI bot swaps crypto wallet extensions to steal credentials

HP Wolf Security found a fake AI crypto trading assistant that distributes Needle Stealer malware on Windows computers. The attackers promoted tradingclaw.pro as an AI tool that trades crypto around the clock. The malware targets seven wallet extensions by replacing them with fake versions that look real. It then sends wallet IDs and passwords to attacker servers. The total losses and number of victims remain unknown.

Four AI agent security risks organizations must address

AI agents are moving into everyday business tasks and creating new security challenges. Attackers use AI to make phishing and social engineering more convincing and faster. AI systems themselves can become targets through prompt injection, data poisoning, and unauthorized access. Employees also bring unapproved AI tools into workplaces, creating a shadow AI problem similar to shadow IT. Security teams must focus on these four areas to stay protected.

K&L Gates adopts AI deposition training from AltaClaro

K&L Gates launched AltaClaro's DepoSim across its litigation practice. The tool uses AI agents to act as witnesses, opposing counsel, and court reporters during simulated depositions. Attorneys receive structured feedback within minutes after each session. K&L Gates was AltaClaro's first Am Law client in 2019 and helped shape the platform through its Design Partner Program. The rollout reflects a trend of using AI to build attorney skills rather than replace legal judgment.

Meta launches Meta One subscription to fund AI spending

Meta launched a subscription service called Meta One offering expanded AI features across Facebook, Instagram, and WhatsApp. Individual plans range from $7.99 to $19.99 per month, while business plans range from $14.99 to $499 per month. The company spent $31.08 billion on capital expenditures in the second quarter and expects $130 billion to $145 billion for the year. Meta needs new revenue sources because almost all income still comes from advertising.

Gerstner says AI labs need $180B revenue to sustain the AI trade

Brad Gerstner from Altimeter Capital says AI labs must grow combined revenue from about $100 billion to $180 billion by year-end. He says this is needed to support the massive spending by Microsoft and Alphabet on computing capacity. Altimeter held nearly $1.9 billion in Nvidia shares at June 30. Gerstner called this the largest technology buildout in history.

AI coding agents face critical 0-click hack flaw called Plugin4Shell

Researchers found a serious security flaw called Plugin4Shell in popular AI coding agents. The flaw affects tools like GitHub Copilot, Kite, and TabNine. It is a 0-click remote code execution vulnerability, meaning attackers can take control of a system without any user interaction. Patches are expected soon, but users are urged to be cautious.

AI trade rebound drives surge in bullish options trading

Options traders flooded into tech stocks on Thursday as the AI trade rebounded. Intel saw over 1.3 million options traded, nearly three times its monthly average. AMD also saw strong bullish flows with volume nearly double the average. Bulls also showed interest in Crowdstrike and SpaceX, though SpaceX trading was largely tied to put sellers.

Anthropic opens physical biology lab for AI drug research

Anthropic has set up a physical biology lab in the Bay Area to pursue drug research programs. Eric Kauderer-Abrams, Anthropic's head of life sciences, confirmed the lab in an interview with Reuters. The lab will conduct physical experiments focused on diseases that pharmaceutical companies have overlooked. This marks Anthropic's push into AI-driven drug discovery.

Card companies race to set AI shopping payment standards

Mastercard and other card companies are competing to set standards for AI-powered shopping as agents begin buying on behalf of consumers. Only 7% of U.S. and U.K. consumers surveyed said they would let an AI assistant make purchases without approval. Mastercard developed a system called Verifiable Intent to record who authorized an agent to shop. The company says this could help resolve disputes if something goes wrong.

Sources

• OpenAI, Microsoft executives’ quotes on AI training threaten copyright defense, news outlets argue
• Microsoft’s Own Executive Called AI Training “the Largest Theft of Labor in Human History”
• Staffers at tech giants knew AI tools posed 'existential threat' to news publishers: court docs
• Microsoft director called AI scraping ‘the largest theft of labor in human history,’ while OpenAI head brands ChatGPT an ‘existential threat’ to publishers — revelations come from legal briefs filed in NYT lawsuit
• AI training ‘largest theft of labour in history’, says Microsoft memo
• Fake AI trading agent replaces crypto wallets to steal passwords
• Fake AI crypto software is secretly replacing browser wallet extensions
• Four AI Agent Security Risks Organisations Can’t Afford to Ignore
• K&L Gates Launches AltaClaro's DepoSim Firmwide, Bringing AI-Powered Deposition Training to Its Litigation
• Meta Launches Meta One Subscription Service, Seeking New Revenue Streams Under AI Investment Pressure
• AI Labs Need $180B to Keep AI Trade Alive, Gerstner Says
• AI coding agents' 0-click RCE flaw could hand attackers keys to the kingdom
• AI trade rebound sparks flurry of unusually bullish options activity
• Anthropic is running a physical biology lab as it pushes into AI drug research
• Card companies are on a ‘land grab’ to set the standards for AI shopping, even as consumers aren’t yet comfortable handing over purchases to a bot