---
date: 2026-05-19
subject: "Pope Leo on AI and human dignity | Commencement speakers draw AI boos | OpenAI beats Musk suit"
---

**Pope Leo XIV will release his first encyclical**, "Magnifica Humanitas," on May 25 alongside Anthropic co-founder Christopher Olah, addressing human dignity, labor and warfare in the AI era. **Public backlash against AI is outpacing the industry's growth,** with former Google CEO Eric Schmidt becoming the third commencement speaker booed in three weeks. **An Oakland federal jury rejected Elon Musk's suit** against OpenAI, Sam Altman and Microsoft as time-barred, removing a hurdle to the company's for-profit conversion and potential $1 trillion IPO. **AIPN convenes a panel Thursday on AGI national security implications** featuring Ray Kurzweil, Anthropic national security policy head Tarun Chhabra, and Mark Beall. Senators Mark Warner and Mike Rounds will deliver opening remarks.

# 1. AI Policy Today

- **Pope Leo XIV to release first encyclical "Magnifica Humanitas" on AI and human dignity May 25 with Anthropic co-founder** — Pope Leo XIV will present "Magnifica Humanitas" at the Vatican on May 25, 2026 alongside Anthropic co-founder Christopher Olah and theologians Anna Rowlands and Léocadie Lushombo, the Vatican said Monday. The Chicago-born pontiff signed the document on May 15, 135 years to the day after Leo XIII signed "Rerum Novarum" on the Industrial Revolution. The Vatican said the encyclical addresses "the protection of the human person in the age of artificial intelligence" and is expected to consider AI's effects on workers' rights and warfare. [The Guardian](https://www.theguardian.com/world/2026/may/18/pope-leo-encyclical-human-dignity-ai-anthropic)

- **Eric Schmidt becomes third commencement speaker booed in three weeks as AI sentiment polls low** — Public backlash against AI is outpacing the industry's growth, the Wall Street Journal reported Tuesday, citing "booed commencement speakers, blocked data centers, plummeting poll numbers." Former Google CEO Eric Schmidt drew boos at the University of Arizona on Friday after telling graduates AI "will touch every profession, every classroom, every hospital, every laboratory." It was the third such commencement reaction this month, after speeches at the University of Central Florida and Middle Tennessee State University, SiliconANGLE reported. Semafor cited polling showing 70% of Americans think AI is moving too fast and 18% of young people feel hopeful about it, with Gen Z workers more than three times as likely as older workers to say AI's risks outweigh its benefits. Schmidt's family office Hillspire has invested in more than a dozen AI startups between 2019 and 2025, including Anthropic. [WSJ](https://www.wsj.com/tech/ai/the-american-rebellion-against-ai-is-gaining-steam-94b72529?mod=rss_Technology) [SiliconANGLE](https://siliconangle.com/2026/05/18/eric-schmidt-booed-commencement-speech-ai-remarks/) [Semafor](https://www.semafor.com/article/05/18/2026/ai-skepticism-grows-among-us-youth)

- **Federal jury in Oakland rejects Musk's claims against OpenAI as time-barred, clearing path to roughly $1 trillion IPO** — A nine-member federal jury in Oakland on Monday rejected Elon Musk's lawsuit against OpenAI, Sam Altman, Greg Brockman and Microsoft after under two hours of deliberation, finding the claims were barred by the three-year statute of limitations. Judge Yvonne Gonzalez Rogers immediately adopted the advisory verdict and dismissed the case. OpenAI lawyer William Savitt told reporters jurors viewed the lawsuit as an "after-the-fact contrivance" by Musk to sabotage a competitor, AP reported. Musk's attorneys signaled they are weighing appeal, and Musk wrote on X that the verdict creates "a precedent to loot charities," Axios reported. The dismissal removes a major legal obstacle to OpenAI's for-profit conversion as the company, currently valued at about $852 billion, moves toward a potential public offering at roughly $1 trillion. [WSJ](https://www.wsj.com/tech/ai/jury-sides-with-openai-sam-altman-in-case-brought-by-elon-musk-933240ff?mod=rss_Technology) [AP](https://www.apnews.com/article/0b9b0bfaffe96f2c930341f52dfe4f8c) [Axios](https://www.axios.com/2026/05/19/musk-altman-openai-trial)

- **Senate panel Thursday on AGI national security implications with Anthropic policy head, Kurzweil and Mark Beall** — On Thursday, Senators Mark Warner and Mike Rounds will deliver opening remarks at an AI Policy Network-convened panel on the national security implications of artificial general intelligence. The panel will feature AI futurist Ray Kurzweil, AI Policy Network president Mark Beall, and Tarun Chhabra, Anthropic's head of national security policy and a distinguished visiting fellow at the Hoover Institution. The session will cover recent AI technical milestones including Project Mythos and pathways to AGI, emerging national security capabilities from cyberwarfare to the battlefield, and a flexible policy response framework to prepare the U.S. government for AGI risks while pressing American competitive advantage in AI development. [Luma](https://luma.com/zoi4v9im)

- **Anduril VP confirms on-record use of Anthropic's Claude in defense prototypes despite Pentagon supply chain risk designation** — Anduril vice president of mission systems Quay Barnett said on the record that the company is testing Google Gemini, Meta Llama and Anthropic Claude in defense prototypes, MIT Technology Review reported in a profile published Monday. The disclosure surfaces Anthropic's Claude in active Pentagon adjacent testing even as Anthropic remains under its federal supply chain risk designation. The profile details Anduril and Meta's prototyping of AR glasses for military helmets under the Army's $159 million Soldier Borne Mission Command (SBMC) prototype contract, with Anduril's self-funded EagleEye helmet/headset combo announced in October running on a parallel track. The Army is not expected to move SBMC into production until 2028. [MIT Technology Review](https://www.technologyreview.com/2026/05/18/1137412/inside-anduril-and-metas-quest-to-make-smart-glasses-for-warfare/)

- **NextEra to acquire Dominion in $67 billion all-stock deal** — NextEra Energy announced Monday it will acquire Dominion Energy in an all-stock transaction valued at about $67 billion, combining two major utilities that will serve roughly 10 million customers across Florida, Virginia, North Carolina and South Carolina, WSJ reported. Dominion shareholders will receive a fixed 0.8138 NextEra shares per Dominion share plus a one-time $360 million cash payment at closing, while NextEra stockholders will own 74.5% of the combined business, AP reported. The deal is expected to close in 12 to 18 months and requires approval from both companies' shareholders, the Nuclear Regulatory Commission and federal and state regulators. NextEra and Dominion said they would offer Dominion customers in Virginia and the Carolinas a total of $2.25 billion in bill credits over two years, CBS News reported, though Clean Virginia warned the headline figure is a one-time payout rather than ongoing relief. Dominion powers hundreds of data centers across Virginia, and the combined company will operate dual headquarters in Juno Beach, Florida and Richmond. [WSJ](https://www.wsj.com/business/energy-oil/nextera-to-buy-dominion-energy-in-67-billion-deal-cc3c94b0?mod=pls_whats_news_us_business_f) [AP](https://apnews.com/article/1916dc2187883c0d4eaf69ce11c51c75) [CBS News](https://www.cbsnews.com/amp/news/nextera-dominion-67-billion-deal-ai-data-centers/)

- **Nvidia CEO Jensen Huang predicts China market will reopen to U.S. AI chips "over time," days after Trump-Xi summit** — Nvidia CEO Jensen Huang said Monday he believes China's market will reopen to U.S. AI chip suppliers, telling Bloomberg Television in an interview at a Dell event that "my sense is that over time, the market will open," Investing.com reported. Huang accompanied President Trump on the May 14-15 Beijing trip with Xi Jinping, which produced no immediate breakthrough for Nvidia to sell H200 chips. Nvidia has received U.S. government licenses to sell its H200 chip to China but lacks Chinese approval, with Beijing backing domestic chip suppliers. The Trump administration excluded the more advanced Blackwell and Rubin chips from January's limited H200 export approval. Huang has previously described the China market as a $50 billion opportunity for Nvidia. [Investing.com](https://www.investing.com/news/stock-market-news/nvidia-ceo-says-he-believes-china-market-will-open-over-time-4696684)

# 2. China Watch

- **CXMT files for Shanghai STAR Market IPO with state fund backing** — China's leading DRAM memory chip maker ChangXin Memory Technologies, known as CXMT, filed for a Shanghai STAR Market IPO with first quarter 2026 revenue up about 700% year over year, Pandaily reported. The cap table is anchored by central government industry funds and Guangdong linked investors holding roughly 5%. Memory output is expanding alongside the domestic AI compute buildout, including high-bandwidth memory used in training and inference clusters. The listing extends Beijing's state capital push into indigenous DRAM and HBM supply, a layer of the AI stack targeted by U.S. Commerce Department export rules restricting advanced node memory shipments to China. [Pandaily](https://pandaily.com/cxmt-star-market-ipo-guangdong-capital-2026)

- **Hygon's C86-5G server CPU approaches 5GHz as Intel supply crunch opens a domestic window** — Chinese chip designer Hygon's next generation C86-5G server CPU is on the cusp of 5GHz clock speeds, with up to 128 cores, four-way SMT and AVX-512 instructions, Pandaily reported. Intel supply constraints and price hikes have created a widening opportunity for Hygon and other domestic chipmakers in the Chinese data center market. The chip widens memory bandwidth with DDR5-5600 support across 16 channels, narrowing the gap with current generation Intel Xeon parts. Hygon has been on the U.S. Commerce Department's Entity List since 2019, and the C86 roadmap signals one route by which Chinese AI clusters can substitute domestically designed x86 silicon for sanctioned U.S. server CPUs. [Pandaily](https://pandaily.com/chinese-server-cpu-5ghz-hygon-c86-4g-2026)

- **XPeng rolls out China's first mass-produced robotaxi powered by in-house Turing chips** — XPeng's first mass-produced robotaxi rolled off the line in Guangzhou, the company's first full-stack self-developed L4 autonomous vehicle, Pandaily reported. The vehicle carries four of XPeng's in-house Turing AI chips delivering 3,000 TOPS of onboard computing and runs a pure vision driving model without LiDAR or high-definition maps. Pilot operations are slated for the second half of 2026, with fully autonomous service targeted for early 2027. The launch advances an all-domestic Chinese AV stack from chip to autonomy software, hardening Beijing's L4 timeline as the U.S. Commerce Department's connected vehicle rule keeps Chinese-origin automated driving systems out of the U.S. market. [Pandaily](https://pandaily.com/xpeng-robotaxi-production-vehicle-2026)

- **China Telecom launches token subscription plans, treating AI compute as a carrier billed utility** — State owned China Telecom rolled out trial commercial AI token subscription plans on World Telecom Day, May 17, with personal plans starting at 9.9 yuan ($1.40) per month for 10 million tokens, Pandaily reported. Developer and small-business tiers run from 39.9 yuan ($5.50) up to 299.9 yuan ($42) per month. The carrier is shifting its commercial model from selling cellular data to metering generative AI inference for retail consumers. The packaging puts AI inference on China's state controlled telecom billing rail, a distribution channel that gives Beijing direct visibility into per user AI consumption. [Pandaily](https://pandaily.com/china-telecom-token-subscription-plans-may2026)

# 4. Capability & Research Watch

- **Forcepoint details supply chain attack that turned LiteLLM into a multicloud credential stealer** — Forcepoint X-Labs attributed a supply chain compromise of LiteLLM, the unified gateway library serving more than 100 LLM providers, to a threat group it tracks as TeamPCP, the security firm published Monday. Two malicious LiteLLM versions, 1.82.7 and 1.82.8, were pushed to PyPI after attackers pivoted through a poisoned Trivy build pipeline to exfiltrate a PYPI_PUBLISH token from LiteLLM's CI/CD runner. The payload scanned compromised systems for OpenAI, Anthropic, Microsoft Azure, AWS and Google Cloud API keys, plus kubeconfig and AWS credential files, and exfiltrated them via curl to a lookalike domain at models.litellm.cloud. Persistence ran through a Sysmon.py polling backdoor that contacted checkmarx.zone every 50 minutes. [SiliconANGLE](https://siliconangle.com/2026/05/18/forcepoint-details-teampcp-supply-chain-attack-turned-litellm-credential-stealer/)

- **Anthropic relaxes Mythos confidentiality, letting Project Glasswing partners share vulnerability findings externally** — Anthropic revised its earlier position to allow users of its Mythos cybersecurity model to share information about cyber threats with other partners and outside organizations, Investing.com reported. Mythos is being deployed under Project Glasswing, a controlled program covering Amazon, Microsoft, Nvidia and Apple. Partners may now disclose their involvement and at their discretion share findings, best practices, tools or code developed under the program. Anthropic said partners may share that information with security teams at other companies, industry bodies, regulators and government agencies, open source maintainers, the media or the public, subject to responsible disclosure norms. The Pentagon is deploying Mythos to find and patch software vulnerabilities across the U.S. government even as it races to complete a transition away from Anthropic, the Defense Department's top technology official said last week. [Investing.com](https://m.investing.com/news/stock-market-news/anthropic-to-let-partners-share-mythos-cybersecurity-findings-with-others-4697228?ampMode=1)

# 5. Industry & Market Watch

- **Standard Chartered to cut 7,800 back-office jobs by 2030 in AI driven consolidation** — Standard Chartered said it will cut 15% of its corporate function roles by 2030, about 7,800 redundancies out of more than 52,000 back-office staff, citing AI as a driver of slimmer operations and higher profitability, The Guardian reported. Most affected centers are in Chennai, Bengaluru, Kuala Lumpur and Warsaw. The cuts reflect automation and AI adoption, with some staff reskilling. CEO Bill Winters said the cuts are "replacing in some cases lower-value human capital with the financial capital and the investment capital we're putting in." The bank's total global workforce stands at roughly 82,000, and StanChart set aside $190 million in precautionary provisions linked to the Iran conflict in the quarter. [The Guardian](https://www.theguardian.com/business/2026/may/19/standard-chartered-bank-cut-jobs-ai-london)

- **Blackstone to invest $5 billion in Google TPU powered AI cloud venture targeting 500MW in 2027** — Blackstone will invest an initial $5 billion in equity in an AI cloud venture with Alphabet's Google, becoming majority owner and planning to bring 500 megawatts of data center capacity online in 2027, CNBC reported. The U.S. based venture will offer Google's custom Tensor Processing Units, or TPUs, through a compute-as-a-service model to compete with CoreWeave in the AI cloud market. Total investment value could reach $25 billion including leverage, Investing.com reported, citing Bloomberg. Blackstone appointed Benjamin Sloss, a longtime Google executive, as CEO of the new venture, while Google Cloud chief executive Thomas Kurian said the venture would address growing TPU demand by offering additional ways to access computing capacity. [CNBC](https://www.cnbc.com/2026/05/19/blackstone-google-ai-data-center-joint-venture-tpu.html) [Investing.com](https://www.investing.com/news/stock-market-news/google-blackstone-to-create-new-ai-cloud-company-wsj-reports-4697320)

- **Anthropic acquires SDK toolmaker Stainless and will wind down hosted products for rival labs** — Anthropic acquired Stainless, the New York based SDK generator company founded by former Stripe engineer Alex Rattray and used by OpenAI, Google, Cloudflare, Replicate and Runway, TechCrunch reported Monday. Anthropic told TechCrunch it will wind down all hosted Stainless products including the SDK generator, though existing customers retain full rights to modify and extend SDKs they have already generated. The Information reported last week that the deal was in talks at more than $300 million; Stainless was backed by Sequoia Capital and Andreessen Horowitz. Stainless software has powered every official Anthropic SDK since the earliest days of its API. [TechCrunch](https://techcrunch.com/2026/05/18/anthropic-has-acquired-the-dev-tools-startup-used-by-openai-google-and-cloudflare/)

- **Meta moves 7,000 workers into AI roles ahead of 10% Wednesday layoff** — Meta is reassigning 7,000 workers to new initiatives related to AI workflows under an internal memo from Chief People Officer Janelle Gale, ahead of layoffs hitting about 10% of its workforce, or roughly 8,000 employees, Wednesday, Investing.com reported. The combined transfers and cuts will affect about 20% of Meta's workforce, which stood at 77,986 employees at the end of March. Gale wrote that many leaders will announce organizational changes and that orgs will adopt flatter structures with smaller pod-style teams. Meta has also closed an additional 6,000 open roles as part of the process, and employees in North America were told to work from home Wednesday. [Investing.com](https://www.investing.com/news/stock-market-news/exclusivemeta-lays-out-plans-for-may-20-layoffs-restructuring-internal-document-says-4696710)

- **Joint Interagency Task Force 401 awards $500 million counter-UAS contract to Perennial Autonomy** — Joint Interagency Task Force 401 awarded a three-year Indefinite Delivery/Indefinite Quantity contract with a $500 million ceiling to Perennial Autonomy for AI enabled counter-unmanned aerial system operations, the War Department announced Monday. The contract covers Merops interceptors, Bumblebee quadcopters and Hornet midrange strike drones currently employed by U.S. Central Command forces. The systems integrate detection, tracking and engagement using computer vision, radio frequency sensing, jam-resistant communications and autonomous targeting, while retaining warfighter authority over the use of lethal force. Army Brig. Gen. Matt Ross, JIATF 401 director, flew the Bumblebee V2 interceptor at a Fort Bragg exercise on May 14 alongside key joint force leaders. [War Department](https://www.war.gov/News/News-Stories/Article/Article/4495165/joint-interagency-task-force-401-awards-500-million-counter-uas-contract/)

# 6. Global & Geopolitics

- **EU Commission opens 35-day consultation on draft Article 6 guidelines defining high-risk AI systems** — The European Commission today published draft guidelines interpreting Article 6 of the AI Act, the operative provision that determines which AI systems must undergo conformity assessment, registration and post-market monitoring. The text addresses two high-risk pathways: AI systems used as safety components of products covered by Annex I harmonised legislation requiring third-party conformity assessment, and AI systems falling within the use cases listed in Annex III. The Commission opened a parallel targeted consultation running May 19 to 10 p.m. CET on June 23, 2026, restricted to responses submitted through an anonymous online questionnaire. Stakeholders can browse the draft through a guidelines explorer on the AI Act Single Information Platform. [European Commission](https://digital-strategy.ec.europa.eu/en/library/draft-commission-guidelines-classification-high-risk-ai-systems) [European Commission](https://digital-strategy.ec.europa.eu/en/consultations/targeted-consultation-draft-guidelines-classification-high-risk-artificial-intelligence-systems)

- **Ofcom to update UK codes to force AI deepfake removal, urges hash-matching deployment** — UK media regulator Ofcom said Monday it will change its codes of practice to force service providers to detect and remove intimate image abuse and AI generated deepfakes, with the new code expected to come into force in autumn 2026 subject to parliamentary process, The Guardian reported. Ofcom is urging sites to deploy hash-matching technology to automatically block recirculation of violating imagery. The move follows a threatened legal challenge from campaign group End Violence Against Women, whose lawyers said Ofcom was failing its obligations to protect women and girls. Prime Minister Keir Starmer said in February that deepfake nudes and "revenge porn" must be removed within 48 hours or technology firms risked being blocked in the UK. [The Guardian](https://www.theguardian.com/technology/2026/may/18/tech-firms-uk-rules-intimate-image-abuse)

- **Helsing and OHB form KIRK joint venture for AI enabled tactical reconnaissance satellites** — Germany's defense AI startup Helsing and space company OHB launched a joint venture called KIRK to develop a space based tactical surveillance, reconnaissance and target acquisition system using AI for near real-time targeting, OHB said in a German language EQS-News release Tuesday. Helsing and OHB will jointly lead an expanded consortium with Kongsberg Defence & Aerospace and HENSOLDT that originally formed in December 2025, with OHB joining as a new partner. The consortium will use software-defined satellites and AI optimized onboard capabilities to compress latency between data capture and target engagement. Helsing co-CEO and co-founder Gundbert Scherf cited the war in Ukraine as evidence Europe must field integrated space based defense systems quickly to maintain orbital sovereignty. [onvista](https://www.onvista.de/news/2026/05-19-eqs-news-helsing-und-ohb-gruenden-joint-venture-kirk-fuer-taktische-aufklaerung-aus-dem-weltraum-0-37-26513265)

- **Iran conflict tightens AI chip supply chain through PCB shortages, 40-week semiconductor lead times** — Geopolitical risk from the Iran conflict is compounding AI driven demand and tightening availability across printed circuit boards, semiconductors, optical interconnects and power components, Data Center Knowledge reported. PCB prices jumped as much as 40% in April as manufacturers grappled with shortages of copper foil and epoxy resin. Victory Giant Technology, a Chinese PCB supplier to Nvidia, warned Middle East conflict could further increase copper costs, which already account for roughly 60% of PCB raw material expenses. Supply chain intelligence firm Accuris reported semiconductor lead times reached roughly 40 weeks in March 2026, a shift it described as moving from creep to crisis as shortages spread across the electronics market. CNBC reported this morning that the chip sector is scrambling to shore up access to key materials as the Iran conflict raises costs even as stocks continue to rally amid the AI boom. [Data Center Knowledge](https://www.datacenterknowledge.com/supply-chain/iran-conflict-threatens-pcb-supply-chain-as-ai-server-demand-surges) [CNBC](https://www.cnbc.com/2026/05/19/iran-war-ai-chip-supply-chain-costs.html)

- **New Zealand's Deepfake Digital Harm and Exploitation Bill heads to first reading this week** — New Zealand's Deepfake Digital Harm and Exploitation Bill is set to go through first reading this week with support across the political spectrum, criminalizing the creation, sharing or sale of sexually explicit deepfakes without consent, per an op-ed by University of Canterbury law lecturer Cassandra Mudgway carried by NZ City News. The amendment responds to the spread of Grok generated sexualized material, with reports of up to 3 million sexualized images circulating. The legislation places New Zealand alongside the UK, Australia, South Korea and the U.S. in introducing or expanding deepfake criminalization laws. Mudgway writes that banning the apps used to produce such material should be the next legislative step. [NZ City](https://home.nzcity.co.nz/news/article.aspx?id=446161&fm=psp,tsf)