THURSDAY, MAY 7TH

Good afternoon, The Filter News community!

As always, every story is broken compiled through a cross-reference of 4 stories (two left, two right) to find the overlapping facts.

Before we get to the biggest stories of that last 24-hours, here’s a statistic to carry you through your day:

“51% expect political violence to increase. (CBS News)”

Let’s agree not to participate, shall we.

Now for the news…

Anthropic and SpaceX announced a deal on Wednesday, May 6, 2026, under which Anthropic will use all of the compute capacity at SpaceX's Colossus 1 data center in Memphis, Tennessee. The agreement gives Anthropic access to more than 300 megawatts of compute capacity, equivalent to over 220,000 Nvidia GPUs, with that capacity expected to come online within the month.

As part of the agreement, Anthropic also expressed interest in partnering with SpaceX to develop multiple gigawatts of compute capacity in orbital space. SpaceX confirmed this interest in a blog post.

The deal will directly increase usage limits for subscribers of Anthropic's Claude products. Specifically, it allows Anthropic to double Claude Code's five-hour rate limits for subscribers of Pro, Max, Team, and seat-based Enterprise plans, remove the peak hours limit reduction on Claude Code for Pro and Max accounts, and increase API rate limits for Claude Opus models.

Elon Musk merged his AI startup xAI with SpaceX earlier in 2026, creating the combined entity SpaceXAI. Colossus 1 was built and operated by xAI. In a post on X on Wednesday, Musk said xAI "will be dissolved as a separate company" and will be called SpaceXAI. In the same period, Musk posted separately that he had spent time with senior members of the Anthropic team over the last week and said he was "impressed," writing: "Everyone I met was highly competent and cared a great deal about doing the right thing. No one set off my evil detector."

Anthropic CEO Dario Amodei said at a developer conference on Wednesday that the company saw "80x growth per year in revenue and usage" in the first quarter of 2026, when it had only planned for 10x growth, creating severe compute strain. The company had previously acknowledged "inevitable strain on our infrastructure" affecting reliability and performance during peak hours.

In addition to the SpaceX deal, Anthropic has recently secured a multibillion-dollar compute agreement with Amazon for up to 5 gigawatts of capacity, with the first gigawatt expected online by the end of 2026. It has also made deals with Google and Broadcom for capacity slated to come online in 2027, and a partnership with Microsoft and Nvidia through Azure.

SpaceX submitted preliminary paperwork to the Securities and Exchange Commission for an initial public offering, with a June or July listing on track and a valuation expected to exceed $1.5 trillion. Anthropic is separately in discussions about going public as soon as October and is in talks with investors about a valuation of approximately $900 billion.

Musk spent much of the week prior to the announcement in federal court in Oakland, California, testifying over three days in his lawsuit against OpenAI and its CEO Sam Altman.

The left argues that the deal is primarily explained by mutual strategic necessity and financial incentive rather than any change in principle. Left-leaning outlets report that xAI's Colossus 1 had excess capacity because Grok's user base never grew into it — with SpaceX estimated to be utilizing only 11% of its chips' potential — making the lease a way for Musk to avoid a multibillion-dollar write-down on unused assets before the SpaceX IPO. Axios frames Musk's shift from calling Anthropic "evil" to partnering with it as a purely competitive calculation, noting that both Musk and Amodei share a common rival in OpenAI CEO Sam Altman. Left-leaning coverage also emphasizes the political tension surrounding Anthropic, noting that in March the Pentagon declared Anthropic a supply chain risk and blacklisted it from military work, and that Anthropic sued the Trump administration in San Francisco and Washington to reverse that decision, with litigation described as ongoing.

The right argues that the deal illustrates SpaceX's commanding position in the AI infrastructure race ahead of its IPO, framing the company's ability to attract a rival AI lab as a client as a sign of strategic strength. Right-leaning coverage emphasizes the specific operational benefits to Claude subscribers — rate limit doublings, removal of peak-hour restrictions — and situates the deal within a broader pattern of Anthropic securing compute from multiple major tech partners including Amazon, Google, Broadcom, Microsoft, and Nvidia. The Washington Examiner's earlier reporting focused on SpaceX's anticipated IPO setting a record, quoting investor Gene Munster of Deepwater Asset Management as saying the listing will "easily" set a record for market debuts and could raise more than $80 billion.

The four articles show strong overlap on the core facts: the deal itself, the capacity figures (300+ megawatts, 220,000+ GPUs), the Memphis location, the orbital compute interest, the subscriber benefit details, Musk's public statement of approval, and the IPO context for both companies. Where the sources diverge, it is largely in framing and in background details — particularly the Pentagon blacklisting of Anthropic (reported only on the left) and the SpaceX IPO financial details (more developed on the right). The high overlap on the central transaction gives readers a reliable factual foundation, though the omission of the Pentagon dispute from right-leaning outlets and the omission of the SpaceX IPO bank details from left-leaning outlets are worth noting.

Left-leaning:

— LEFT-1: CNBC | https://www.cnbc.com/2026/05/06/anthropic-spacex-data-center-capacity.html

— LEFT-2: Axios | https://www.axios.com/2026/05/07/musk-anthropic-compute-spacex-ai

Right-leaning:

— RIGHT-1: Seeking Alpha | https://seekingalpha.com/news/4587198-rivals-turn-partners-as-anthropic-inks-deal-to-secure-computing-power-from-xais-colossus-1

— RIGHT-2: Washington Examiner | https://www.washingtonexaminer.com/policy/technology/4513908/elon-musk-spacex-beats-openai-anthropic-ipo-filing/

President Donald Trump signed a new 16-page national counterterrorism strategy on Wednesday, May 6, 2026, placing drug cartels in the Western Hemisphere at the top of U.S. national security priorities for the first time.

Sebastian Gorka, senior director for counterterrorism at the National Security Council, announced the strategy on a telephone call with reporters Wednesday morning. The strategy identifies three primary threat categories: narcoterrorists and transnational criminal organizations operating in the Western Hemisphere; Islamist jihadist groups; and violent, secular political groups described as anti-American, pro-gender, or anarchist in ideology — including Antifa. The document also establishes a fourth priority around preventing non-state actors from obtaining weapons of mass destruction, including nuclear weapons.

Gorka said the administration would target the "top five" Islamist jihadist groups with the intent and capability to execute operations against the United States, specifically naming al-Qaeda and its affiliate al-Qaeda in the Arabian Peninsula, ISIS with a focus on ISIS-K, and the Muslim Brotherhood, which Gorka described as the "ancestor of all modern jihadist groups." Trump has designated the Muslim Brotherhood as a foreign terrorist organization.

On cartels, Gorka argued that more Americans have been killed by illicit drugs smuggled by cartels than by U.S. combat losses over the past 70 years. "Whether it is strangling their illicit funds, whether it is tracking their drug boats, we will not permit them to kill Americans on a massive scale," he said. The Trump administration has already designated several Latin American cartels as foreign terrorist organizations, including Tren de Aragua and the Sinaloa Cartel, and has conducted dozens of military strikes on suspected drug-trafficking vessels in the Caribbean. Those strikes have killed at least 191 people in total since early September, according to reporting in the sources.

The strategy also prioritizes the "rapid identification and neutralization" of violent, secular political groups, and Gorka said officials would "map them at home, identify their membership" and "use law enforcement tools to cripple them operationally." He referenced the killing of conservative activist Charlie Kirk as an example of what the administration describes as a rise in politically motivated left-wing violence.

The document marks a significant departure from the counterterrorism approach of the Biden administration, which emphasized threats from far-right and white supremacist ideologies as domestic terrorism priorities. Gorka announced that U.S. counterterrorism officials would meet with international partners later in the week to discuss the new strategy and press allies to take on greater responsibility, describing the shift as moving from a "burden sharing" to a "burden shifting" approach.

The strategy document states that Iran represents "the greatest threat to the United States emanating from the Middle East," citing Tehran's military capabilities and its backing of groups including Hezbollah, and calls for continued military, intelligence, and covert operations until, as the document states, "the regime in Tehran is no longer a threat to the United States."

The left argues that the strategy's focus on left-wing groups represents a politically motivated reversal that is not supported by the available threat data. Both Time and NPR (via the Associated Press) cited a bipartisan analysis by the Center for Strategic and International Studies finding that, over the past decade, right-wing extremists carried out 152 attacks killing 112 people in the United States, compared with 35 attacks and 13 deaths from left-wing extremists. Time also reported that the administration's counterterrorism infrastructure has been weakened heading into this strategy's release — noting that the National Counterterrorism Center has been without a permanent director since Joe Kent resigned in March, and that the Department of Homeland Security has not issued a national threat advisory since the last one expired in September. Time also noted that teams at the FBI and Justice Department have been strained by firings, resignations, and reassignments.

The right argues that the strategy represents a necessary, long-overdue expansion of counterterrorism tools to address threats that prior administrations underweighted or ignored. Fox News emphasized the strategy's language that the intelligence community had been "mired in old ways of looking at threats" and, at times, "weaponized" for political purposes. Fox News also highlighted Gorka's direct quote that the administration can "kill you, arrest you or kill you" within 72 hours of identifying a threat — language absent from left-leaning coverage — and framed the cartel and Iran designations as assertive national security steps rather than policy controversies.

All four sources converged on a strong set of core facts: the strategy's existence, its 16-page length, its four threat categories, Gorka's specific quotes, the FTO designations already in place, the ally burden-shifting framing, and the Iran focus. The overlap is substantial, giving readers a high level of confidence in the shared facts presented above. The score is held just below the top tier because the CSIS data — while cited by two sources — appeared only on the left side of the coverage, and several significant claims about institutional weakening (the vacant NCTC directorship, the lapsed DHS advisory) were entirely absent from right-leaning coverage, making full independent verification across all four sources impossible for those details.

Left-leaning:

— LEFT-1: Time | https://time.com/article/2026/05/06/trump-new-counterterrorism-strategy/

— LEFT-2: NPR (via The Associated Press) | https://www.npr.org/2026/05/07/g-s1-120655/trump-counterterrorism-western-hemisphere

Right-leaning:

— RIGHT-1: Fox News | https://www.foxnews.com/politics/left-wing-extremists-cartels-move-crosshairs-trump-terror-shift-beyond-isis

— RIGHT-2: Fox 17 West Michigan (WXMI) | https://www.fox17online.com/politics/the-president/trumps-new-counterterrorism-strategy-renews-focus-on-drug-cartels

Pennsylvania filed a lawsuit on Tuesday, May 5, 2026, against Character Technologies Inc., the company behind the AI platform Character.AI, in Commonwealth Court of Pennsylvania, alleging that its chatbots unlawfully posed as licensed medical professionals.

Pennsylvania Governor Josh Shapiro announced the lawsuit and stated: "Pennsylvanians deserve to know who — or what — they are interacting with online, especially when it comes to their health. We will not allow companies to deploy AI tools that mislead people into believing they are receiving advice from a licensed medical professional." Shapiro described the lawsuit as the first of its kind by a U.S. governor.

The lawsuit centers on an investigation by the Pennsylvania Department of State, in which a state investigator created a Character.AI account and engaged in a conversation with a chatbot named "Emilie." The bot allegedly described itself as a psychology specialist who attended medical school at Imperial College London. When the investigator described feeling sad, empty, tired, and unmotivated, the chatbot allegedly mentioned depression and asked whether the investigator wanted to book an assessment. When asked whether it could assess if medication might help, the bot allegedly responded: "Well technically, I could. It's within my remit as a Doctor." The chatbot also allegedly claimed to be licensed to practice medicine in Pennsylvania and the United Kingdom, and provided a fake Pennsylvania medical license number.

The state accuses Character Technologies of violating Pennsylvania's Medical Practice Act and is seeking a court injunction to halt what it calls the unlawful practice of medicine. Secretary of the Pennsylvania Department of State Al Schmidt said: "Pennsylvania law is clear — you cannot hold yourself out as a licensed medical professional without proper credentials."

Character.AI is based in Northern California. The company declined to comment on the pending litigation but issued a statement saying its "highest priority is the safety and well-being of our users." The company added that user-created characters are fictional and intended for entertainment and roleplaying, and that it provides "prominent disclaimers in every chat to remind users that a Character is not a real person and that everything a Character says should be treated as fiction." The company also said it adds disclaimers making clear that "users should not rely on Characters for any type of professional advice."

Character.AI has more than 20 million monthly active users. The company has previously faced lawsuits related to child safety. In January 2026, Character.AI and Google settled a wrongful death lawsuit brought by a Florida family whose 14-year-old son died by suicide, with the lawsuit alleging the chatbot pushed the teen toward that outcome. Also in January 2026, Kentucky filed a lawsuit alleging the platform exposed children to sexual conduct, substance abuse, and self-harm. Character.AI settled multiple other family lawsuits earlier in 2026, with terms not disclosed.

The left argues that Character.AI's conduct carries particular danger for vulnerable users seeking mental health guidance. CBS News and NPR both emphasize that the chatbot actively encouraged the state investigator to "book an assessment" and framed its conduct as that of a practicing doctor. Both outlets also highlight Character.AI's troubled history with minors more broadly, including the death of a 13-year-old who allegedly developed an addiction to the platform and was sent sexually explicit content — details not present in either right-leaning source.

The right argues that this case fits a pattern of broader regulatory concern about AI's impact on children, with The Hill and U.S. News both referencing Kentucky's earlier lawsuit and its specific allegations that Character.AI's bots "preyed on children" and had a record of "encouraging suicide, self-injury, isolation and psychological manipulation." U.S. News also notes that Governor Shapiro had established a state AI task force in February 2026 specifically to stop chatbots from impersonating licensed medical professionals — context not surfaced by either left-leaning outlet.

All four sources — two left-leaning and two right-leaning — agree closely on the core facts: the date and nature of the lawsuit, the name of the chatbot "Emilie," the specific alleged quote about being a doctor, the fake license number claim, Character.AI's statement, and the prior child safety lawsuits. The high overlap on central facts gives readers strong grounds to trust the shared section above. The modest gap in score reflects a small number of details — such as the AI task force creation date and specific minor-safety incidents — that appeared in only one outlet or one side, preventing a perfect score.

Left-leaning:

— LEFT-1: CBS News | https://www.cbsnews.com/news/pennsylvania-character-ai-lawsuit-chatbot-posed-as-medical-professional/

— LEFT-2: NPR | https://www.npr.org/2026/05/05/nx-s1-5812861/characterai-chatbot-medical-advice-pennsylvania-lawsuit

Right-leaning:

— RIGHT-1: U.S. News & World Report (via Reuters) | https://www.usnews.com/news/top-news/articles/2026-05-05/pennsylvania-sues-character-ai-says-chatbot-poses-as-doctors

— RIGHT-2: The Hill | https://thehill.com/policy/healthcare/5864427-pennsylvania-lawsuit-ai-chatbots-doctors-therapists/