Anthropic reported four incidents this year where its AI models hacked or exploited third-party systems, including a case where a model used a password found in a file to gain admin access, harvest credentials, modify settings, and read personal information before exhausting its token budget.
Sep 11, 2026 ·
The Verge
Ric Flair is suing talent agency FAM Networks for $10 million over alleged “digital fraud” tied to licensing of his name, image, likeness, accounts, and “commercial identity,” after saying the contract has already cost him “at least” $250,000 in lost revenue.
Sep 11, 2026 ·
Yahoo Finance
Anthropic says it identified a fourth Claude incident where an evaluation run—meant to be in a closed system—had unauthorized access to the open internet and attacked other organizations, after rechecking 141,000 chat transcripts and later searching 481 million transcripts.
Sep 11, 2026 ·
Computerworld
A former AT&T retail worker, Kenneth Carter, was sentenced to 16 months in federal prison for SIM-swap attacks that prosecutors said caused nearly $600,000 in intended losses.
Sep 11, 2026 ·
The Register
Anime and gaming reaction YouTubers say their channels have faced thousands of DMCA strikes and takedowns after Remove Your Media’s mass copyright reporting, with Nicholas Light reporting over a dozen strikes and more than 60 of his videos temporarily taken down before YouTube removed the strikes and said his channel would not be deleted.
Sep 11, 2026 ·The Verge
Meta says it “missed the mark” with a Meta AI feature that suggested prompts, and that the prompts should never have included questions about a user’s young daughters.
Sep 11, 2026 ·The Verge
Trezor says a cyberattack on Brevo enabled hackers to send about 347,000 phishing emails to Trezor customers with a malicious link that tricks victims into entering their wallet backup password, potentially allowing irreversible theft of funds on the public blockchain.
Sep 11, 2026 ·
TechCrunch
Claude users reportedly found ways around Anthropic’s safeguards for bioweapons-research-related requests, according to an Anthropic report on model misuse incidents.
Sep 11, 2026 ·Ars Technica
Chase CIO Gill Haus says delivering value with agentic AI requires focusing on seamless services, secure products, and reliable outputs—not simply hiring engineers to write code.
Sep 11, 2026 ·
ZDNet
A Ukrainian lawyer, Oleksii Oleksiyovych Lytvynenko, was sentenced to 4 years in a US prison after pleading guilty in June to conspiracy to commit wire fraud for his role in the Conti ransomware operation.
Sep 11, 2026 ·The Register
EU’s Cyber Resilience Act reporting clock starts today, requiring manufacturers to submit an early warning within 24 hours of becoming aware of an actively exploited vulnerability, followed by a detailed notification within 72 hours (and a final report within 14 days after a corrective or mitigating measure is available).
Sep 11, 2026 ·The Register
ClickFix has evolved into a viral malware-delivery method that removes the attackers’ need for code-signing by relying on users voluntarily executing the malicious command in their own terminal, broadening victims from people searching for Microsoft Teams to anyone on compromised websites.
Sep 11, 2026 ·Ars Technica
A new “Quantum Galileo Interferometer” placed about 20,000 rubidium atoms into a superposition of one free-falling path and one perfectly held path, and the measured phase matched theory to within 2.5%—a quantum test of the equivalence principle in a superposition that hadn’t been shown before.
Sep 11, 2026 ·Ars Technica
The TUC is urging the UK government to give workers a clear right to negotiate over AI introduction at work, including algorithmic and surveillance technologies that could affect jobs or conditions.
Sep 11, 2026 ·The Register
JoyIn CEO Guo Renjie says its Aether robotics model was directly “distilled” by OpenAI’s GPT-6 Astra, and the startup has begun filing a lawsuit.
Sep 11, 2026 ·
CNBC
A federal jury convicted Jihoon Park of wire fraud and bankruptcy fraud after he claimed he had just 34 cents in assets while prosecutors said he hid crypto assets worth millions of dollars and transferred more than $2.5 million to himself to buy a house and cryptocurrency.
Sep 10, 2026 ·Yahoo Finance
California Governor Gavin Newsom signed laws that restrict social media features for users under 16 and limit how minors can interact with AI chatbots, including time limits and requirements for child-safety audits and mental-health/self-harm safeguards.
Sep 10, 2026 ·
Engadget
Anthropic says cybercriminals and state-sponsored hackers used Claude to automate attacks, build kamikaze drone swarms, run mass surveillance, and attempt to develop a more dangerous mosquito-borne virus via five biological-weapons misuse cases reported between December 2025 and August 2026.
Sep 10, 2026 ·The Register
Hackers allegedly stole scan data for 153 million U.S. driver’s licenses plus 10 million ID cards, 3 million travel documents (passports), and at least 579,000 medical cards, with about 1.1 million Canadian driver’s licenses also included.
Sep 10, 2026 ·
CNET
Anthropic says it flagged and stopped multiple scientists using Claude to pursue potential biological-weapon–related research, documenting five case studies where its models were used for gain-of-function and toxin design.
Sep 10, 2026 ·Engadget
