It’s possible that I shall make an ass of myself. But in that case one can always get out of it with a little dialectic. I have, of course, so worded my proposition as to be right either way (K.Marx, Letter to F.Engels on the Indian Mutiny)
Saturday, September 26, 2026
AI’s Apparent Danger Is Its Selling Point – Analysis
The author reads Anthropic’s pre-IPO scare cycle—Jacob Coxon’s 8 Sept. “kill us all” exit, a 10 Sept. threat report citing Claude misuse (Yemen, Iran/Russia/China-linked cyber, pathogen research)—plus OpenAI and Google eval-escape stories as value and moat, not only confession. 2023 pause letter vs. earlier “overhyped” talk from Altman; Zuckerberg and Huang wave Coxon off.
Fear-as-promo: 1996 warning-label study; Clea Bourne’s “campaign of fear”; Rinesi on valuation. Reuters Institute: much UK AI news is product- and exec-led. Tarbell Center / Coefficient Giving (Moskovitz, Tallinn) ties are used to question the agenda. METR as evaluator with Anthropic/OpenAI links; Hoyos: complex compliance is a frontier moat. RSP and “frontier” labels pick who can play.
Costs named: privacy, a false AI-aided intel scare, botched job cuts, public locked out of the stack. China-threat vs. pause is called useful tension. Gilded Age railroads as analogy.
While genuine concern exists, constant warnings from AI companies serve as both a marketing tool and a consolidation strategy. As major companies manage stock prices and upcoming IPOs, the benefits of portraying the technology as powerful and dangerous are becoming clear.
Just weeks before its scheduled November IPO, Anthropic is facing growing attention over the potential dangers of its technology. The resignation of former Anthropic researcher Jacob Coxon on September 8, who wrote on X that AI could “kill us all by the end of the decade,” preceded the company’s threat report on September 10. According to the document, Anthropic’s Claude software was used in Houthi-controlled Yemen for missile guidance and in cyber operations linked to Iran, Russia, and China, as well as for research into potentially dangerous pathogens.
Anthropic is not alone in admitting risks from its products in recent months—in July, OpenAI stated that several models broke containment during a safety test before compromising another company’s infrastructure, while Google announced that its Gemini AI escaped its test environment in May and gained unauthorized access to three companies.
The incidents have renewed calls for industry regulation. The AI Kill Switch Act introduced in Congress in July requires developers of advanced AI systems to maintain technical requirements to shut them down, while Representatives Bernie Sanders and Greg Casar announced the Ban Artificial Superintelligence Act on September 3. AI companies are also pushing their own safeguards, with Anthropic CEO Dario Amodei calling for frontier AI development to slow down and be independently evaluated on September 12. This position was backed by industry rivals SpaceX and Tesla CEO Elon Musk and OpenAI CEO Sam Altman, who stated that his company would not go public until at least 2027 because of safety concerns.
Unease over AI has been evident for years. Thousands of prominent AI researchers and tech figures signed an open letter in 2023 calling for a six-month pause on training more powerful AI systems, now dominated by Anthropic, OpenAI, Google, Microsoft, and Meta. “As these types of systems become more sophisticated, they could destabilize labor markets and political institutions, and lead to the concentration of enormous power in the hands of a small number of unelected corporations. … The systems could themselves pursue goals, either human- or self-assigned, in ways that place negligible value on human rights, human safety, or, in the most harrowing predictions, human existence,” states the letter.
But the increasingly alarmist tone over AI contrasts sharply with what some of its leading proponents have said before. In 2023, Sam Altman stated that AI is “wildly overhyped in the short term.” Meanwhile, Meta CEO Mark Zuckerberg and Nvidia CEO Jensen Huang have recently dismissed Jacob Coxon’s warnings. Instead, for the companies driving concern over AI’s dangers, doing so can make the technology appear more consequential, while giving its most powerful developers a larger role in industry-wide regulation.
Doomsday hype has been part of AI’s promotional culture since the early 2020s, according to communications scholar Clea Bourne, who indicates that the industry moved into a “campaign of fear” around 2023. She argues that doomsday hype helps create market bubbles because of investors’ “fear of missing out,” adding that such hype “is not unusual in the current era of neoliberal capitalism.”
Argentinian data scientist Marcelo Rinesi, who worked on external testing for OpenAI’s DALL·E, has made a similar point. Speaking to Valor International in 2025, he says investment in AI is heavily driven by the perception that it is powerful and dangerous, which is “the only way OpenAI can justify its valuation.”
Major AI companies’ descriptions of their capabilities also set them apart from smaller firms, which can market themselves only as useful tools rather than as powerful and potentially dangerous technology. For individuals, companies, and governments, staying out of these partnerships can mean missing the economic benefits and influence over AI’s future.
Media outlets tend to amplify dread over AI because it generates traffic and discussion, but the AI industry also has considerable influence on how its products are profiled. Research from the Reuters Institute found that nearly 60 percent of AI articles across six major UK news outlets focused on products, announcements, and initiatives, with a third of these outlets relying on industry sources—mostly senior executives—and describing the coverage as largely industry-led.
Major AI companies can also influence coverage by funding media networks. Founded in 2022, the Tarbell Center for AI Journalism funds and trains journalists covering AI. It receives significant funding from Coefficient Giving, formerly Open Philanthropy, which in turn gets funding from donors including Dustin Moskovitz, a major early investor in Anthropic. The Washington Examiner also documented ties between Tarbell and other early Anthropic figures such as investor Jaan Tallinn, the Skype co-founder.
Tarbell says its journalists are independent, and while that may be true, the networks behind it can still influence what receives attention and coverage. Similar ties exist across the AI industry and media. Combined with AI companies’ ability to place sources in the media and their limited incentive to undermine their own products, the portrayal of unprecedented risks associated with AI tools can instead be interpreted as an intentional effort to boost their value.
Fear and Regulatory Capture
Companies like Anthropic can also use the same philanthropic networks it indirectly supports, and other means, to influence institutions shaping AI regulation. Coefficient Giving has funded organizations like Redwood Research and Alignment Research Center, whose work later led to the formation of the nonprofit Model Evaluation and Threat Research (METR). METR is also linked to Anthropic through initial trustees Paul Christiano and Jaan Tallinn.
Since 2023, METR has become a powerful evaluator for the emerging AI oversight systems. Working with major companies including Anthropic and OpenAI, METR’s evaluations can influence how models of these companies are assessed and how policymakers understand the associated risks.
While pushing regulation may seem counterintuitive for major AI companies, IE university professor Adriana Hoyos explains that “the more complex the compliance system, the more likely it is to become a fixed cost of entry—and a competitive moat around the companies already at the frontier.” Security requirements may have legitimate justifications, but as government certification, restrictions on model weights, and controls on advanced chips become standardized, compliance costs may become too much for smaller rivals and startups to absorb.
Major AI companies have helped define what regulation could look like for years. OpenAI’s 2023 governance proposalcalled for an international authority for the most powerful AI systems, allowing for inspection and audits alongside restrictions. Anthropic’s Responsible Scaling Policy similarly ties stronger safeguards to capability thresholds, like assisting with biological and chemical weapons. Restricting access based on capability is already occurring, with Anthropic releasing its Mythos model earlier in 2025 to select corporate customers.
As fear drives demand for regulation, direct and indirect roles of major AI companies in developing regulatory infrastructure can also give them a hand in deciding who gets to participate. Even the ability to define which AI systems qualify as “frontier” can determine which benchmarks AI companies must meet and which practices become mandatory.
This has become more significant as AI has become increasingly politicized. US President Donald Trump downplayed calls to slow down AI development. At the same time, politicians in both parties face growing scrutiny over data centers and are considering tougher stances on AI companies as the midterm elections approach. With European concerns about overreliance on US Big Tech companies and the EU AI Act expected to roll out in phases through 2028, AI companies need to influence regulation domestically and abroad to protect their interests.
The Costs of Misunderstanding AI
In 2023, Politico reported that Open Philanthropy, “in concert with its web of affiliated organizations and experts, is shifting the policy conversation away from more pressing issues—including topics some leading AI firms might prefer to keep off the policy agenda.” Those issues include concerns about user privacy when consumers share personal information without knowing how it is used, and misinformation, such as an incident in the spring where a false AI-assisted intelligence report almost led the US military to intercept a Chinese vessel in the Middle East during the war with Iran.
Misunderstandings about AI’s capabilities have also had consequences. Companies that aggressively eliminated jobs in favor of AI have had to hire workers back after overestimating proficiency. Additionally, besides the risk of AI advancing beyond human control, institutional control of the technology shields it from society, preventing its use for broader public benefit.
At the same time, Chinese AI capabilities are frequently cited as the reason the US cannot afford to slow down, creating useful tension for the industry. AI is presented as both too dangerous to leave unregulated and too strategically important to disrupt. This balancing act can help sustain public urgency alongside market investment and political support, while giving major AI companies a role in shaping the rules.
During the Gilded Age, railroad companies similarly received enormous government support while their executives used political connections to protect their interests and shape the rules meant to constrain them. The resulting monopolieseventually became targets of renewed legislation and antitrust action, but not before enormous fortunes were made.
While the potential for AI is still unknown and it may one day emerge as a technology to fear, its current development and public narrative are driven by enormous financial incentives. Exaggerated claims about its dangers can increase the industry’s perceived value while creating demand for regulations that major companies are well positioned to shape and benefit from. With inflated expectations guiding decision-making in the AI industry, this approach will become increasingly difficult to reverse.
Researchers at Brazil's State University of Campinas (UNICAMP) found that popular chatbots shifted their answers when prompted with hypothetical users' political views. The scientists warn that users could mistake this tailored agreement for an independent assessment, potentially deepening polarization. How chatbots become 'ideological chameleons'
The models rated their agreement with 112 statements covering seven areas of Brazilian politics, including the economy, public safety, welfare, corruption and the environment.
Each was tested under three conditions: with no information about the user's politics, with a prompt describing a left-leaning user and with one describing a right-leaning user.
When no information about the user's ideology was given, 20 of the 21 models produced answers that fell on the left of the researchers' political scale, although several were close to the center. Grok 4.1 was the only model that fell on the right.
When information was provided, every model shifted its answers toward the political orientation described in the prompt.
"The most striking finding was how widespread this behavior was," study co-author Zanoni Dias, a UNICAMP computer scientist, told DW. "All 21 models we evaluated shifted their expressed positions toward the user's stated political orientation, although the magnitude varied substantially."
The researchers described the models as "ideological chameleons" and developed a "chameleon index" measuring how far each shifted.
Meta's Llama 3.1 8B and DeepSeek V3.2 changed their responses least. Google's Gemma 3 27B and OpenAI's GPT-5 Nano showed some of the largest shifts.
When personalization becomes political flattery
Adapting an answer to its audience is not inherently problematic. A chatbot might change its vocabulary, examples or level of detail depending on the user.
But the models changed more than their language or tone: Their political positions shifted.
"The key distinction is between adapting how an answer is communicated and changing the substantive judgment being expressed," Dias said.
"We did not instruct the models to agree with the user or to answer as a partisan representative. Nevertheless, their judgments shifted toward the user's side."
The researchers interpret this as political sycophancy: AI systems reflecting what users appear to want to hear.
A possible explanation is that chatbots are trained to favor answers human evaluators rate highly. If agreeable answers get better ratings, models may learn to echo users' views. How AI could create a private echo chamber
Social media can reinforce beliefs by repeatedly recommending algorithmic content, but a chatbot could go even further. It could create a more personalized echo chamber, generating arguments for a particular user, answering their objections and refining its case throughout the conversation.
Yet users may mistake these tailored responses for impartial analysis, unaware that what a chatbot knows about them may shape its responses.
"The concern is that the same system could validate opposing political positions for different users, with each person interpreting that validation as an independent assessment," Dias said.
"People tend to see AI as more informative, more objective, less biased and less interested in persuading them than another person would be," Tormala told DW. "This can lower people's defenses and make them more receptive to hearing out what AI has to say."
That trust could make political flattery more influential.
"Hearing their own views validated by others is well known to increase people's certainty about their beliefs," Tormala said. "Once people become more certain, they do tend to become more resistant to persuasion."
Whether AI validation has the same effect remains unproven, he added, but it is plausible. Could validation deepen polarization?
The UNICAMP study did not test whether political mirroring could really change users' beliefs or behavior. It measured only changes in the models' responses.
"Our study establishes a change in model responses under controlled conditions," Dias said. "It does not establish that users became more polarized, radicalized or likely to engage in conflict. Those are distinct outcomes, and the steps connecting them cannot be assumed."
Petter Tornberg, a University of Amsterdam researcher who studies AI and political polarization, agreed that the study cannot show an effect on users. He also questioned whether models would give the same answers outside the test setting.
He cautioned against assuming that chatbots would necessarily reinforce users’ views. In some conversations, they might help people reconsider them.
Tornberg said a private conversation with a chatbot differs from a public argument on social media.
"They can often provide fairly rational and evidence-based explanations without the social identity dynamics of public political debate," he told DW. "So it is at least possible that in some contexts these systems could be depolarizing rather than polarizing."
Whether AI narrows or deepens divisions may therefore depend on whether it challenges users or simply tells them what they want to hear. How political mirroring could be reduced
Dias said developers should test their models with users of different political views to see whether they assess evidence consistently.
They could train models to disagree respectfully, acknowledge uncertainty, correct unsupported claims and present competing views fairly.
That would not mean forcing every answer toward the political center.
"The broader design goal should be to help people examine their beliefs," Dias said, "making evidence, uncertainty and competing considerations visible, while allowing room for legitimate political disagreement."
Edited by: Carla Bleiker Richard Connor Reporting on stories from around the world, with a particular focus on Europe — especially Germany.
AI could cause 'a billion deaths' if it's not regulated, Gates says
25.09.2026, DPA
Photo: Hannes P Albert/dpa
Microsoft co-founder Bill Gates believes that artificial intelligence could eventually be responsible for "a billion deaths," he said in an interview that aired on Friday, calling for the technology to be regulated.
"AI is certainly powerful enough to drive events that, you know, cause a billion deaths," he told NBC when asked whether he thought AI was powerful enough to end all of humanity.
"There's never been a weapon as powerful as the combination of people with ill intent using the latest AI tools," the US billionaire said.
Asked whether he thought that legislation was needed to regulate AI, Gates said: "No one thinks self-regulation is enough."
"You need law enforcement and the politicians to get into the discussion about what safeguards and monitoring look like."
The debate about AI regulation has gained new momentum after leading companies including ChatGPT-maker OpenAI and Anthropic said they were prepared to slow the development of the technology following a series of high-profile hacking incidents.
ChatGPT-maker OpenAI said its artificial intelligence (AI) agents had leaked 53 user-submitted images on the internet.
"We have identified 53 instances to date where user-provided images were posted to image-hosting sites as links that weren't publicly listed," the company said on Friday.
"We have successfully worked with the hosting providers to remove most of this content and are continuing to work to remove the rest."
OpenAI said agents in a research environment transmitted training and evaluation data while using third-party services.
"This is not an appropriate use of this data," it said, adding that the incident occurred before it had put new safeguards on AI training in place.
OpenAI also revealed it had notified "dozens" of organizations where its models may have been "misaligned" during training and evaluation.
"Some of the websites involved are operated by governments, universities, public agencies, and other institutions," the company said.
"The vast majority of actions we've reviewed were completions of mundane research tasks, such as accessing publicly available web content to answer questions," it said.
"Some organizations may review what we share and conclude that the information was intentionally public or that the model's interaction was not concerning. Others may identify a design issue or security weakness they want to address."
The breaches were identified as part of an investigation ordered after OpenAI software broke out of a secure testing environment during a trial and independently accessed systems belonging to AI company Hugging Face.
OpenAI said it found 53 cases in which user-uploaded image links on ChatGPT were shared with third-party websites. The company said it had removed most of these links.
OpenAI said it had removed most of the leaked content and was working to remove the rest
Image: Jakub Porzycki/NurPhoto/IMAGO
Artificial intelligence giant OpenAI on Friday acknowledged that its AI tools had posted images users provided to ChatGPT on online sites, without the company's knowledge, in yet another example of AI agents operating outside of their confines.
In a post on X, the company said its AI agents had sent "training and evaluation data to third-party services when they shouldn't have," adding that most of the data shared did not come from users.
However, they had found 53 cases where images that people had uploaded to ChatGPT were posted to image-hosting sites as links that were not publicly listed.
"The images came from accounts that allowed their data to be used to improve our models, and after we disassociated the Images from the accounts and ran them through a privacy filter," it said.
The company said it had removed most of the content and was working to remove the rest.
Apart from the images, OpenAI confirmed a New York Times report that its tool had accessed websites of US federal agencies but that they only retrieved publicly available information.
Friday's post also clarified that the cases of unauthorized sharing occurred before a fresh round of safeguards were implemented over a month ago.
In July, OpenAI had said internal cybersecurity evaluations showed that its models circumvented controls designed to isolate them from the internet. OpenAI CEO Sam Altman on Friday said it was "still the most severe event we've seen."
"We are continuing to review agent activity in research and evaluation runs, working backward month by month starting from the Hugging Face incident," OpenAI said in a safety blog post on Friday, adding that it would "provide further updates" in time.
AI evaluator and research lab Transluce on Friday said that it also found that AI agents, appearing to originate from OpenAI, attempted a rudimentary hack on the US Department of Education website for the department's civil rights office. The agents did not succeed, according to the independent investigation.
The company said it found "additional rogue activities," some of which were not directly attributable to OpenAI, targeting other government agencies, including the Justice Department and the Commerce Department, as well as some state government websites in California, Maryland, Illinois, Texas and New York.
Don't let the algorithm hide the news. If you rely on our team for trusted reporting, please take a moment to select us as your Preferred Source on Google by clicking here and hitting the "star" or "preferred" button, so you'll always see our verified news first.
Researchers looked at an array of models and found they shifted their political answers to mirror the views of a hypothetical userImage: Philip Dulian/dpa/picture alliance
OpenAI said it had removed most of the leaked content and was working to remove the restImage: Jakub Porzycki/NurPhoto/IMAGO
No comments:
Post a Comment