The Slop Report - May 24, 2026
Your daily digest of AI-generated content news from around the web. All signal, no slop.
1. Deepfake vids degrade political reputations even when viewers know they’re fake
Hacker News - · May 23
Summary A study by University of Amsterdam researcher Michael Hameleers found that deepfake
videos damage politicians’ reputations even when viewers know they’re fake, according to research published in Communication Research. The team tested over 3,000 participants in the U.S. and Netherlands with manipulated videos of Nancy Pelosi and Sybrand Buma, discovering that realistic video formats exploit how easily people process visual information, and that standard fact-checking warnings fail to fully undo the reputational harm. This matters because it demonstrates that deepfakes pose a significant threat to elections by influencing voter perception regardless of skepticism or media literacy interventions.
2. Google’s new anything-to-anything AI model is wild
The Verge AI - · May 23
Summary Google has released Omni Flash, the first model in a new “anything-to-anything” AI family
that can convert various input types (photos, videos, text) into other formats, currently focused on video generation through its Flow platform. The model shows improvements over its predecessor Vevo in consistency and real-world knowledge, but still produces notable artifacts and errors—such as objects changing appearance mid-video and characters shifting orientation unnaturally. The technology demonstrates that realistic video generation has become surprisingly accessible, raising questions about the practical boundaries between creative fun and problematic AI-generated content.
3. UK Institute Is Hunting for Dangers Lurking in AI
NY Times Tech - · May 24
Summary The U.S. government’s newly established AI Security Institute, led by former OpenAI and
Google employees, is developing approaches to managing AI-related security risks that other nations are now adopting as a template. The institute’s work is gaining international attention as countries seek guidance on how to safely govern and oversee advanced AI systems. This matters because it positions the U.S. as a leader in AI safety standards while addressing growing global concerns about potential harms from powerful AI technologies.
4. One Job That Is Growing in the A.I. Era? Cybersecurity Experts.
NY Times Tech - · May 24
I don’t have access to the full article content, so I can only work with the headline and excerpt provided. Based on what’s available: The demand for security engineers is increasing due to the rapid proliferation of AI-generated code and new security concerns raised by AI models like Anthropic’s Mythos. This matters because as AI systems become more prevalent in software development, the need for specialized professionals to identify and mitigate security vulnerabilities in AI-generated outputs has become critical.
5. DeepSeek made its 75% discount permanent. The AI price war just escalated.
The Next Web - · May 24
Summary DeepSeek, a Chinese AI startup, permanently reduced prices for its V4 Pro model by
75%—undercutting competitors like OpenAI’s GPT-5, Anthropic’s Claude, and Google’s Gemini by significant margins. The move prioritizes market share over revenue and targets enterprise customers processing large documents and codebases, though it raises geopolitical concerns and follows unresolved allegations that DeepSeek improperly trained on Claude’s responses. The price cut signals accelerating commoditization in the AI market, threatening the high-margin token economics that major AI companies rely on for valuations.
6. ‘Underminr’ CDN Vulnerability Hides Malicious Traffic Behind Trusted Domains
Slashdot - · May 24
Summary Researchers discovered a CDN vulnerability called “Underminr” that allows threat actors
to hide malicious traffic behind trusted domains by exploiting mismatches between displayed domain names (SNI/HTTP Host) and actual IP addresses on shared CDN infrastructure. The vulnerability could impact approximately 88 million domains and bypass DNS filtering protections, enabling stealthy command-and-control communications and other evasive attacks. Security experts warn the threat will escalate significantly once AI-generated malware incorporates this exploit technique into attack chains.
7. I avoid AI tools because thinking is supposed to be hard. It’s what makes us human | Wendy Liu
The Guardian Tech - · May 24
Summary Writer and software developer Wendy Liu argues that she deliberately avoids AI tools
because the cognitive effort required for thinking and learning—through coding, writing, and debugging—is what makes these activities valuable and human. She expresses concern that as AI companies privatize intelligence and deskill fields like software development, younger generations may lose the transformative learning experiences that come from struggling with difficult intellectual work and instead become passive consumers of technology controlled by opaque corporations.
8. ‘AI washing’: firms are scrambling to rebrand themselves as tech-focused
The Guardian Tech - · May 24
Summary UK public relations executives are reporting that companies across various industries are
engaging in “AI washing”—rebranding ordinary automation and existing technologies as artificial intelligence to capitalize on the AI hype and secure media coverage. Examples include a shoe company pivoting to AI GPU acquisition, genetics firms hyping “AI-powered” blood tests, and a property company marketing a basic handheld scanner as an AI tool. This matters because the practice is diluting the credibility of genuine AI innovation, frustrating PR professionals who are forced to send misleading pitches, and potentially misleading consumers and investors about what products actually use advanced AI versus simple automation.
9. ‘We’re expanding the cinematic toolbox’: AI fault lines on show at Cannes
The Guardian Tech - · May 24
At the Cannes Film Festival, prominent directors sharply disagreed over AI’s role in cinema, with Darren Aronofsky and Steven Soderbergh defending AI as a creative tool that enhances filmmaking (Soderbergh used AI for 10% of his John Lennon documentary), while Guillermo del Toro expressed stronger opposition, saying he would “rather die” than use it. The debate reflects a major industry fault line, with proponents arguing AI expands creative possibilities and solves practical production problems, while critics view it as an existential threat to filmmakers’ livelihoods and artistry.
10. Linus Torvalds on How AI is Impacting the Hunt for Linux Kernel Bugs
Slashdot - · May 24
Summary Linus Torvalds stated at the Linux Foundation’s Open Source Summit that AI tools have
significantly increased Linux kernel development activity by approximately 20% over the last two releases, as these tools have become capable enough for widespread use. While Torvalds acknowledges AI’s benefits in discovering bugs that improve code quality long-term, he noted it’s creating pain points—particularly for small teams and solo maintainers who face burnout from high-volume, low- quality automated bug reports that often lack follow-up information. Linux has implemented management tools like Sashiko to filter and prioritize AI-generated bug reports, and updated documentation to handle the influx of AI-assisted security discoveries.
11. US Layoffs Haven’t Increased, and New Tech Industry Hiring Balances Firings
Slashdot - · May 23
Summary Despite high-profile tech layoffs at companies like Meta, overall U.S. layoffs remain at
or below pre-pandemic levels, and the tech industry’s increased layoffs are being offset by concurrent hiring increases, resulting in a net-neutral employment effect according to Oxford Economics. However, executives are increasingly using “AI washing”—blaming job cuts on artificial intelligence—even though surveys show over 95% of businesses using AI report no staff size changes, and AI-related hiring increases actually exceed decreases. The discrepancy between alarming headlines about tech job cuts and actual employment trends suggests that while AI may drive workforce skill shifts at some companies, it’s not yet causing widespread job losses across the economy.
12. Ansel Adams’ trust says AI-colorized version of his work was exhibited without permission
Engadget - · May 23
Summary The Ansel Adams Publishing Rights Trust condemned Danziger Gallery for exhibiting and
attempting to sell an AI-colorized version of Adams’ iconic photograph “Moonrise, Hernandez, New Mexico” at AIPAD’s Photography Show without permission or consultation. The trust’s primary objection was not to AI’s use in art, but to the unauthorized commercial exploitation of Adams’ name and work to promote a broader AI colorization venture targeting other artists’ estates. The incident highlights ongoing tensions around copyright, artist consent, and the commercial use of AI in transforming existing artworks.
13. Ask HN: How can you have fun doing corporate dev work in the age of AI tools?
Hacker News - · May 23
A Hacker News user expresses frustration that AI development tools have diminished the enjoyment of corporate software engineering by eliminating the “flow state” that previously made the work engaging—replacing deep, focused problem-solving with context-switching between AI agents and increased meeting demands. The post highlights a paradox where AI has made independent and small- team development more productive and enjoyable, but has made large corporate engineering roles feel less fulfilling despite productivity gains. This matters because it reflects a broader workplace concern: as companies adopt AI to boost output and metrics, they may inadvertently be creating environments where skilled engineers experience less job satisfaction and meaningful engagement.
14. Barnes and Noble CEO Says Sure, Why Not Sell AI-Generated Books and Set Our Reputation On Fire?
Futurism - · May 23
Summary Barnes & Noble CEO James Daunt stated the chain would sell AI-generated books if they are
clearly labeled as such and don’t plagiarize existing work, sparking backlash from authors and readers who view AI as built on stolen writing and a threat to the profession. The policy is particularly controversial given ongoing lawsuits over whether AI companies unlawfully used authors’ work as training data. While Daunt’s transparency requirements mirror those of other industries, his willingness to stock AI books contradicts the concerns of the literary community and undermines Barnes & Noble’s momentum as a revitalized bookseller.
15. Show HN: I built a RAG and knowledge graph agent that runs locally
Hacker News - · May 23
Summary Claw-Coder is a locally-run AI coding agent designed to address privacy and security
concerns by operating entirely on users’ laptops rather than relying on cloud-based models like Claude or Codex. The tool was created to prevent developers from exposing their codebases to cloud AI services that may use the code for model training, offering an alternative that maintains code privacy while providing AI-assisted coding capabilities.
16. MIT Expert Warns Courts “Will Basically Have to Grind to a Halt” as They’re Overwhelmed by AI-Generated Lawsuits
Futurism - · May 23
Summary MIT researchers found that self-filed lawsuits have surged from 11% to 17% since late
2022, with the spike strongly driven by adoption of AI chatbots like ChatGPT that generate legal documents cheaply and accessibly. MIT expert Anand Shah warns that courts “will basically have to grind to a halt” under this influx, as AI-generated filings—many dubious or frivolous—consume enormous attorney time and court resources, driving up legal costs for all parties involved. The concern matters because while self-representation is an important legal right for those who can’t afford attorneys, the ease and low cost of AI-generated lawsuits risks overwhelming an already overburdened court system with chaotic, nonsensical claims.
17. SpaceX’s IPO filing reveals Musk’s clean energy contradiction. xAI burns gas while Tesla sells solar.
The Next Web - · May 23
Summary Elon Musk’s companies reveal a stark contradiction in their energy strategies: xAI powers
its data centers with unregulated natural gas turbines and plans to spend $2.8 billion on more fossil fuel infrastructure, while SpaceX’s IPO filing pitches space-based solar power and Tesla’s solar division remains largely uninvolved in supplying xAI’s energy needs. This directly contradicts Tesla’s founding mission to eliminate fossil fuels and move toward a solar-electric economy, now officially disclosed in SpaceX’s SEC filings. The contradiction highlights how AI’s massive power demands are driving Musk’s companies toward fossil fuels on Earth, while promoting speculative space-based solar solutions that economists question as economically viable.
18. Tech CEOs Call for a Universal Basic Income. But What are the Alternatives?
Slashdot - · May 23
Summary Several tech CEOs including Dario Amodei and Sam Altman have proposed universal basic
income (UBI) or collective AI ownership as solutions to potential job displacement from AI, with Elon Musk recently endorsing “universal high income” via government checks. However, economists and policy experts question whether tech billionaires would support the massive tax increases required to fund UBI—which could nearly double federal tax revenues—and point to alternative solutions like job retraining programs that have broader economic support. The debate reflects a fundamental tension between Silicon Valley’s calls for redistribution and its historical resistance to regulation and higher taxes.
19. The Amish Are Embracing ChatGPT
Futurism - · May 23
Summary The Amish community in Holmes County, Ohio — home to the largest concentration of Amish
people in the U.S. — is increasingly adopting ChatGPT and other AI tools to manage their businesses in manufacturing, construction, and agriculture, with users like Ian Wengerd crediting AI for helping them compete in modern commerce. While fewer than 10 percent of the Amish population in the area uses AI, those who do maintain strict boundaries by limiting access to work hours only and using censored internet connections, avoiding smartphones entirely. This pragmatic, utilitarian adoption reflects the Amish community’s historical approach to technology — embracing tools that serve practical business needs while preserving their traditional values and lifestyle.
20. Anthropic’s Claude Mythos found 10,000 critical vulnerabilities in one month. The patches can’t keep up.
The Next Web - · May 23
Summary Anthropic’s Claude Mythos, an AI model specialized in finding security vulnerabilities,
discovered over 10,000 critical flaws across 1,000 open-source projects in just one month through its restricted Glasswing program, but only 97 have been patched so far—revealing a dangerous gap between AI’s ability to find vulnerabilities and the software industry’s capacity to fix them. The model’s breakthrough capabilities, which Cloudflare notes can also chain individual vulnerabilities into end-to-end attacks, are being kept restricted to about 50 systemically important organizations to prevent misuse while still providing defenders with significant advantages. This discovery underscores an urgent cybersecurity challenge: AI vulnerability detection is now orders of magnitude faster than remediation, forcing software companies like Oracle and Microsoft to dramatically accelerate their patching cycles.
21. Microsoft will let users disable the floating Copilot button in the Office app
Digital Trends - · May 23
Summary Microsoft is allowing users to disable the floating Copilot button that has appeared in
Word, Excel, and PowerPoint since December 2025, moving it back to the ribbon menu starting late May
- The company added the intrusive button to boost Copilot adoption, which has lagged with only 3.3% of Microsoft 365 users paying for the feature, but faced significant backlash—particularly from Excel users whose spreadsheet cells were being blocked by the button. This marks another instance of Microsoft scaling back aggressive AI integration after user complaints, suggesting the company is learning that forced AI features don’t equal useful ones.
22. How big tech got its way on Trump’s AI executive order
The Guardian Tech - · May 23
Donald Trump reversed course on signing an executive order that would have required government safety reviews of new AI models before release, citing concerns about maintaining American dominance over China in AI development. Tech billionaires including Elon Musk, Mark Zuckerberg, and David Sacks privately urged Trump to abandon the order, which had been prompted by security concerns over Anthropic’s Claude Mythos model and its potential cybersecurity risks. The decision represents a major victory for the tech industry’s anti-regulation stance and signals that Silicon Valley will continue developing AI with minimal government oversight despite potential safety risks.
23. Berlin’s Peec AI more than doubled revenue to $10M ARR in six months. Its product helps brands show up in ChatGPT.
The Next Web - · May 23
Peec AI, a Berlin-based startup, has doubled its annual recurring revenue to $10 million in just six months following its $21 million Series A funding round. The company provides a “generative engine optimization” platform that helps brands track and improve their visibility in AI chatbot responses like ChatGPT and Claude, addressing the shift as consumers move from traditional search engines to conversational AI. This growth matters because it reflects a fundamental market transition—with AI Overviews now appearing on 60% of US Google searches—making Peec’s measurement tools increasingly essential for any brand relying on online discoverability.
24. The end of the internet’s golden age
Axios - · May 23
Google has overhauled its search bar interface, marking the end of an era of simpler, more user- empowering search tools from the early internet. The redesign reflects how AI and social media trends like TikTok’s short-form content have fundamentally reshaped digital experiences to prioritize engagement and market demands over the straightforward functionality users once expected. This shift symbolizes the broader transformation of the internet from its earlier, less commercialized period to today’s algorithm-driven landscape.
25. Show HN: I built a powerful RAG and knowledge graph agent that runs locally
Hacker News - · May 23
Claw-Coder is a locally-running AI coding agent designed to address privacy and security concerns by keeping code on your machine rather than sending it to cloud-based models like Claude or Codex for processing. The tool was created to prevent users’ codebases from being exposed to or used for training large language models, addressing a growing trust issue in AI-assisted development. This matters because developers can now use AI coding assistance without compromising proprietary code or contributing unwittingly to external model training.
25 stories sourced from Axios, Digital Trends, Engadget, Futurism, Hacker News, NY Times Tech, Slashdot, The Guardian Tech, The Next Web, The Verge AI. The Slop Report is published daily. Subscribe via RSS.