Anthropic faces a new lawsuit from the owners of 274 Elvis Presley songs, including hits like Can't Help Falling in Love. The plaintiffs, led by Authentic Brands Group, say the AI company used the catalog to train its Claude model without permission. They are seeking about $41 million in damages, calculated at $150,000 per track. Anthropic said it plans to defend itself against the claims.
The case highlights growing tension over how AI companies train models on copyrighted material. Lead attorney Richard Busch described the scale of the alleged infringement as unprecedented. The lawsuit also claims that Anthropic's guardrails did not prevent the model from producing infringing outputs. It was filed in federal court in San Francisco.
In a separate development, 1Password launched a platform called 1Password for Agents on October 22, 2026. The tool lets AI agents log into accounts without ever seeing a user's actual password. Instead, agents authenticate using a unique digital key. Users must approve each login request on a trusted device before credentials are filled. The company is working on integrations with OpenAI, Anthropic, and Meta, with Grok Bot and Meta Muse among the first connected services.
Microsoft Research open-sourced Agensh, a system that coordinates up to 1,024 AI coding agents without a central leader. Agents claim tasks, share findings, and merge code into a shared repository. Testing on ProgramBench showed improved scores as agent counts grew, reaching a 79.59 percent pass rate with 32 agents. The harness runs on Linux with Docker.
Former Google CEO Eric Schmidt and co-authors compared OpenAI and Anthropic to historic East India Companies in a recent essay. They argue the US and China dominate the global AI market while other nations become dependent. The authors noted US companies plan over $600 billion in investment for 2026, mostly through debt. They also warned that a conflict over Taiwan and chipmaker TSMC poses the greatest threat to the AI supply chain.
Interactive Brokers launched IBKR Autonomous Trading, giving eligible clients a way to connect AI tools that spot opportunities and submit orders. The company also announced a higher quarterly dividend alongside strong third quarter earnings. Meanwhile, SentinelOne introduced Singularity ADR and Autonomous SOC products following U.S. alerts on AI-driven cyber threats. Its stock rose 6.67 percent in one day to reach $24.00.
A new study called MAFT improves AI security testing by sorting data into six semantic groups before analysis. Researchers found that matching scaling methods to data types reduced false alarms from 67.1 percent to 5.6 percent. The improvement came from preprocessing alone without changing the AI model. The work was published in Discover Artificial Intelligence by Sunil Kumar Yadav and colleagues.
Australian voters showed support for local AI rules and data centre safeguards in a Resolve poll of 2,200 people. Respondents wanted Australian-specific laws rather than international agreements and expressed concern about the environmental impact of data centre construction. Reports link new data centres to a potential $50 billion rise in power prices over six years. Legislation is expected early next year covering safety checks and security agency access to frontier models.
Key Takeaways
- Anthropic is being sued over 274 Elvis Presley songs allegedly used to train Claude, with plaintiffs seeking about $41 million in damages.
- 1Password launched 1Password for Agents, letting AI agents log in without seeing actual passwords, with integrations planned for OpenAI, Anthropic, and Meta.
- Microsoft open-sourced Agensh, a multi-agent harness that coordinates up to 1,024 AI coding agents and achieved a 79.59 percent pass rate with 32 agents.
- Eric Schmidt compared OpenAI and Anthropic to colonial trading companies, warning that a Taiwan conflict involving TSMC threatens the AI supply chain.
- Interactive Brokers rolled out IBKR Autonomous Trading, allowing connected AI tools to spot opportunities and submit orders for eligible clients.
- SentinelOne launched Singularity ADR and Autonomous SOC products, with its stock rising 6.67 percent to $24.00 after U.S. alerts on AI-driven cyber threats.
- The MAFT framework reduced AI security false alarms from 67.1 percent to 5.6 percent by sorting data into six semantic groups before analysis.
- Australian voters backed local AI rules in a poll of 2,200 people, with new legislation expected early next year covering safety checks and frontier model access.
- 1Password for Agents is a free upgrade for eligible Individual and Family plans, with Grok Bot and Meta Muse among the first integrations.
- Anthropic said it disagrees with the Elvis lawsuit claims and will fully defend itself against the allegations.
Elvis Presley rights holders sue Anthropic over AI music training
Authentic Brands Group and other rights holders sued Anthropic for using 274 Elvis Presley songs to train its AI model Claude without permission. The songs include hits like Can't Help Falling in Love and All Shook Up. The plaintiffs are seeking $41.1 million in damages. They also want Anthropic to stop using the copyrighted works. Anthropic said it disagrees with the claims and will defend the case.
Elvis Presley catalog owners sue Anthropic for copyright infringement
The estate of Elvis Presley is suing Anthropic for copyright infringement. The company allegedly used 274 Elvis songs to train its AI model Claude without permission. The lawsuit says this use of the music is illegal. The case focuses on how AI systems learn from copyrighted musical works. No damages amount is mentioned in this report.
Anthropic sued for allegedly training AI on 274 Elvis Presley songs
Authentic Brands Group sued Anthropic, saying the AI company infringed on 274 Elvis Presley songs. The lawsuit says Claude was trained on the music from the start. The plaintiffs want $150,000 for each infringed track, totaling about $41 million. Lead attorney Richard Busch called the scale of infringement unprecedented. The suit also says guardrails did not stop the infringing outputs.
Elvis rights holder sues Anthropic over Claude AI training
Owners of 274 Elvis Presley songs sued Anthropic in federal court in San Francisco. They say the AI company used the songs to train its agent Claude without paying. The plaintiffs include Authentic Brands Group and other rights holders. They are seeking about $50 million in damages. Anthropic said it will fully defend itself against the claims.
1Password lets AI agents log in without seeing your password
1Password launched a new platform called 1Password for Agents. It lets AI agents log into accounts without ever seeing the actual password. Users still approve login requests on a trusted device. The company is working with OpenAI, Anthropic and Meta on tighter integrations. Plaintext credentials and one-time codes are not shared with the AI agent.
1Password launches AI login with digital key for agents
1Password launched 1Password for Agents on October 22, 2026. AI agents authenticate using a unique digital key instead of seeing your password. Users must approve access on a trusted device before credentials are filled. The free upgrade covers eligible Individual and Family plans on 1Password.com. Grok Bot and Meta Muse are among the first integrations.
Interactive Brokers rolls out AI autonomous trading
Interactive Brokers Group launched IBKR Autonomous Trading for eligible clients. The tool lets connected AI tools spot opportunities and submit orders. Shares are down 10% over the past month but up 19.41% year to date. The company also announced a higher quarterly dividend and strong third quarter earnings.
Eric Schmidt compares OpenAI and Anthropic to colonial trading companies
Former Google CEO Eric Schmidt and co-authors compared OpenAI and Anthropic to historic East India Companies. The essay argues the US and China dominate the global AI market while other nations depend on them. US companies plan over $600 billion in investment for 2026, mostly through debt. The authors warn that a conflict over Taiwan and chipmaker TSMC is the greatest threat to the AI supply chain.
Four practical tasks to try with AI agents
The article highlights how AI agents like OpenAI's dots and Meta's Muse can handle repetitive tasks. Anthropic is testing a similar agent called Duet for its Claude chatbot. The author tested these tools for two weeks and found they can complete tasks like booking reservations. The piece shows how agents are becoming more flexible and efficient for everyday use.
Australian voters back local AI rules and data centre safeguards
A Resolve poll of 2,200 Australians found voters want Australian-specific AI laws rather than international agreements. Respondents also expressed concern that the data centre construction boom could harm communities and the environment. Reports link new data centres to a potential $50 billion rise in power prices over six years without new energy sources. Legislation is expected early next year covering safety checks, local computing reserves, and security agency access to frontier models.
New Data Method Boosts AI Hack Detection by 23 Points
A study introduces MAFT, a framework that sorts security data into six semantic groups before AI analysis. Researchers found that matching scaling methods to data types cuts false alarms from 67.1 percent to 5.6 percent. The improvement comes from preprocessing alone without changing the AI model. The work was published in Discover Artificial Intelligence by Sunil Kumar Yadav and colleagues.
Microsoft Open-Sources Agensh Multi-Agent Harness
Microsoft Research released Agensh, a system that coordinates up to 1,024 AI coding agents without a central leader. Workers claim tasks, share findings, and merge code into a shared Git repository. Testing on ProgramBench showed scores improved as agent counts grew from 1 to 1,024. The harness runs on Linux with Docker and achieved a 79.59 percent pass rate with 32 agents.
SentinelOne Launches AI Security Tools as Stock Hits $24
SentinelOne unveiled Singularity ADR and Autonomous SOC products following U.S. alerts on AI-driven cyber threats. The stock rose 6.67 percent in one day and 11.21 percent over seven days to reach $24.00. Analysts at Simply Wall St note a fair value of $24.25 but flag a high price-to-sales ratio of 7.6x compared to software peers at 4.1x.
Oppenheimer: Web Software Outlook Improves Amid AI Fears
Oppenheimer reports growing investor interest in web and e-commerce software ahead of Q3 earnings, though AI disruption concerns remain. Shopify is expected to beat revenue forecasts with 33 to 34 percent growth. Wix.com may face slower growth due to shifting demand toward AI tools. GoDaddy and Commerce.com also face mixed outlooks as they adapt to AI-driven market changes.
Sources
- Elvis Presley’s rights holders sue Anthropic saying it trained AI on his music
- Elvis Presley Catalog Owners Sue AI Company Anthropic for Copyright Infringement
- Anthropic sued for allegedly using 274 Elvis Presley songs to train AI models
- Elvis rights holder sues Anthropic, says songs illegally used to train AI
- 1Password will now let AI agents log in to accounts for you without ever needing to see your password
- 1Password AI Login Enhances Security with Digital Key
- Is Interactive Brokers Group (IBKR) Fully Valued On AI Trading Launch And Strong Q3 Earnings?
- Ex-Google CEO Eric Schmidt calls OpenAI and Anthropic modern colonial trading companies
- Put Your A.I. Agents to Work with These 4 Killer Tasks
- 2,200 Aussies Back AI Safeguards: Poll Results [2026]
- Semantic Typing of Security Data Boosts AI Intrusion Detection by Up to 23 Points
- Microsoft Open-Sources Agensh: A Multi-Agent Harness for 1,024 Agents
- SentinelOne (S) Unveiled New AI Security Tools, Is The Stock Now Expensive?
- Oppenheimer sees web, commerce software Q3 backdrop improving, AI worries linger
Comments
Please log in to post a comment.