News

A Federal Judge Ruled the Pentagon Broke the Law Trying to Blacklist Anthropic

English

A federal judge ruled the Pentagon illegally punished Anthropic for refusing to drop safety limits on autonomous weapons, Sony Music and Warner Chappell sued the company for billions over song lyrics used to train Claude, hackers found a way to drain paid Claude accounts without a password, and India's foreign currency reserves hit a record high as Prime Minister Modi opened a new chapter with Central Asia.

The tuput Editors · · 6 min read

Anthropic had a rough week in two courtrooms and a very good one in a third. A federal judge threw out the Pentagon’s attempt to blacklist the company, two of the world’s biggest music publishers sued it for billions, and security researchers caught malware quietly draining people’s paid Claude accounts. Meanwhile India had a quieter, better week: its foreign currency cushion hit an all-time high, and the prime minister spent it opening doors in Central Asia.

The Pentagon tried to punish a company for saying no

Anthropic’s dispute with the Department of Defense started when the Pentagon wanted to use Claude for things Anthropic had built hard limits against: fully autonomous weapons and mass surveillance of American citizens. Anthropic refused to strip those safeguards out. The Pentagon responded by formally labeling the company a “supply chain risk,” a designation normally reserved for security threats, not vendors who won’t agree to a contract’s terms.

Anthropic sued, and this week it won. U.S. District Judge Rita Lin ruled that the designation violated the company’s First Amendment rights and that Anthropic had been denied due process under the Fifth Amendment. Her opinion did not pull punches: “The empty invocation of national security is not a blank check to punish and retaliate against government critics.”

The ruling is especially awkward for the Pentagon because of its own timing. According to the judge’s findings, a day after Defense Secretary Pete Hegseth ordered the designation, a Pentagon official told Anthropic the two sides were “very close” to a deal, and the White House was separately discussing using Anthropic’s newer Mythos model in sensitive settings. Judge Lin found Hegseth ordered the label applied before the department had even finished the analysis meant to justify it.

Two of music’s biggest names came after Claude next

A day after that ruling, Sony Music Publishing and Warner Chappell Music filed their own suit against Anthropic in the same federal court in Northern California, naming CEO Dario Amodei and co-founder Benjamin Mann personally as defendants. The complaint accuses Anthropic of a “brazen campaign” of torrenting, scraping, and downloading copyrighted song lyrics and compositions to train Claude, and names specific tracks including “Eye of the Tiger” and “All I Want for Christmas Is You.”

The publishers say tens of thousands of their works were used without permission. Under US copyright law’s statutory damages rules, that puts Anthropic’s theoretical exposure at several billion dollars: up to 150,000 dollars per work willfully infringed, plus another 25,000 dollars for every instance where the publishers say copyright information was stripped out. They are also asking the court to force Anthropic to hand over an accounting of its training data and destroy infringing copies under court supervision.

This is not Anthropic’s first brush with music and publishing litigation. The company already settled a 1.5 billion dollar suit with book publishers earlier this year, and separately faces copyright claims from Universal Music Group, Concord, ABKCO, BMG, and Round Hill Music. Anthropic says it disagrees with the new claims and will fight them in court.

Your Claude account might not be entirely yours right now

Separately, Anthropic warned this week that infostealer malware already sitting on some users’ computers has been quietly hijacking their logged-in Claude sessions. Security researchers named several malware families behind it, including Vidar, LummaC2, StealC, RedLine, and a Mac-targeting variant called Atomic Stealer, or AMOS.

The trick is that this malware doesn’t need a password or a two-factor code. It steals the browser cookie that proves a session is already authenticated, then replays it to ride along on someone else’s paid account and burn through their usage limits without ever logging in itself. Anthropic says it noticed the pattern because usage allowances kept refilling and then draining while the actual account owners were inactive.

The company is now signing out affected users, stripping saved payment methods from compromised accounts, and refunding charges it identifies as unauthorized. It hasn’t said how many accounts were hit. The advice from security researchers is the same one that applies to any browser-based account: keep antivirus software current, and treat any sudden unexplained usage on an AI subscription as a sign to check for stolen session cookies rather than a billing error.

India’s savings account for hard times just got bigger than ever

India’s foreign exchange reserves climbed to a record 729.33 billion dollars in the week ending August 21, according to Reserve Bank of India data reported this week. That is the eighth straight week of gains, adding roughly 63 billion dollars over two months and pushing past the previous record set back in February.

Much of the jump traces to central bank measures rolled out in June to pull in more foreign currency, including cut-price hedging for state banks and companies borrowing overseas and a free hedging facility for banks raising foreign currency deposits. Between early June and August 21, the Reserve Bank of India logged close to 73 billion dollars flowing in through those schemes alone.

The number matters beyond the headline. That reserve now covers more than eleven months of India’s goods imports and equals roughly 94 percent of the country’s total external debt, a buffer that gives the country real room to absorb a currency shock or a bad year for exports without scrambling.

Modi went looking for new ground in Central Asia

Prime Minister Narendra Modi spent the tail end of this week in Tashkent, on his fourth visit to Uzbekistan, before heading on to Kyrgyzstan for the Shanghai Cooperation Organisation summit. Trade between India and Uzbekistan currently sits close to a billion dollars, modest by the standard of India’s biggest partnerships, and the two governments used the visit to talk about growing it: critical minerals, defense cooperation, healthcare, energy, digital connectivity, and education all featured in the agenda.

It is not a headline-grabbing sum of money yet. What it represents is India methodically building relationships in a part of the world it has historically underinvested in, at a moment when China and Russia both compete hard for influence across Central Asia. A modest trade figure today is the kind of thing that either stays modest or becomes the seed of something bigger a decade from now, and this week’s visit was India betting on the second outcome.

Put the two threads of this week together and a pattern shows up. Anthropic is fighting on three fronts at once and won the biggest one on the merits of a straightforward argument: the government can’t invent a security threat to punish a company for a policy disagreement. India, meanwhile, spent its week doing the unglamorous work of building buffers and relationships that only pay off when nobody’s watching. Neither kind of story trends for long, but both are the kind that actually holds up.

Share
Copied!

Sources & further reading

  1. NPR: Judge says the Pentagon can't designate AI company Anthropic a 'supply chain risk'
  2. NPR: Judge says Pentagon's measures against Anthropic were 'illegal and baseless'
  3. TechCrunch: Anthropic gets its first court win over the Pentagon's supply-chain risk label
  4. NBC News: Federal judge blocks Pentagon blacklisting of Anthropic, calling it 'illegal and baseless'
  5. CNN Business: Judge rules the Pentagon's supply chain risk label for Anthropic unlawful
  6. TechCrunch: Sony Music, Warner sue Anthropic, alleging a 'brazen campaign' of intellectual property theft
  7. Music Business Worldwide: Sony Music Publishing and Warner Chappell sue Anthropic in multi-billion dollar lawsuit
  8. Benzinga: Sony Music, Warner Chappell Sue Anthropic Over Copyright Claims
  9. BleepingComputer: Anthropic warns infostealer malware is hijacking Claude sessions to drain usage
  10. Security Affairs: Infostealers Are Hijacking Claude Sessions and Draining Subscriptions
  11. Bloomberg: India's Forex Reserves Hit Record High on Central Bank Measures
  12. Business Recorder: India's forex reserves hit record $729 billion on sustained capital inflows
  13. Free Press Journal: PM Narendra Modi To Visit Uzbekistan, Kyrgyzstan From Aug 29-Sept 1 To Deepen India's Strategic Ties With Central Asia
  14. India TV News: PM Modi to visit Uzbekistan and Kyrgyzstan, to attend SCO Summit, check details

Researched and written with the help of AI tools and edited for accuracy. Provided for general information and discussion only, not professional advice. See our editorial standards and disclaimer. Spotted an error? Tell us.

#ai#anthropic#pentagon#claude#sony music#warner chappell#copyright#cybersecurity#india#forex reserves#modi#uzbekistan#daily roundup

Enjoyed this? Get the next one.

One good read at a time, straight to your inbox. No spam, unsubscribe anytime.

More in News
Anthropic's New Model Landed the Same Week It Bet $35 Billion on More Computing Power
Anthropic launches a faster Claude alongside a $35 billion cloud deal, Alibaba gives its newest model away for free, and companies are quietly building software instead of buying it. Plus a new power plant in Bihar and an Indian record in T20 cricket.
OpenAI's Newest Model Is So Good at Hacking It Needs a Waiting List
OpenAI's next model can hack well enough that most people won't get to use that part of it, Anthropic reveals a Claude checkpoint trained to hack that passed its own safety audits, and a Madhya Pradesh city finishes clearing five years of trash.
The Pentagon's New AI Chatbot Store Has Room for Everyone Except Claude
The Pentagon adds ChatGPT and Grok to its official AI platform while Claude stays locked out, an AI startup nearly doubles its valuation in nine months, and India's economy grows faster than anyone expected.
← all articles