DeepSeek’s Safety Guardrails Failed Every Test Researchers Threw at Its AI Chatbot
Security researchers tested 50 well-known jailbreaks against DeepSeek’s popular new AI chatbot. It didn’t stop a single one.
Read more at Wired
Topics
-
AI systems could be ‘caused to suffer’ if consciousness achieved, says research
Experts and thinkers signed open letter expressing concern over irresponsible development of technology. Artificial intelligence systems capable of feelings or self-awareness are at risk of being ...The Guardian - 4h -
DeepSeek’s Answers Include Chinese Propaganda, Researchers Say
Since the Chinese company’s chatbot surged in popularity, researchers have documented how its answers reflect China’s view of the world. Some of its responses amplify propaganda Beijing uses to ...The New York Times - 3d -
DeepSeek, ChatGPT, Grok … which is the best AI assistant? We put them to the test
Chatbots we tested can write a mean sonnet and struggled with images of clocks, but vary in willingness to talk politics. ChatGPT and its owners must have hoped it was a hallucination .. But ...The Guardian - 2d -
DeepSeek’s Answers Include Chinese Propaganda, Researchers Say
Since the Chinese company’s chatbot surged in popularity, researchers have documented how its answers reflect China’s view of the world. Some of its responses amplify propaganda Beijing uses to ...The New York Times - 3d -
DeepSeek’s A.I. Chatbot Awkwardly Navigates China’s Censors
Asked about sensitive topics, the bot would begin to answer, then stop and delete its own work. It refused to answer questions like: “Who is Xi Jinping?”The New York Times - 4d -
OpenAI Seems Concerned That DeepSeek Copied Its Work
Did DeepSeek train a model on data it didn’t get permission to use?Inc. - 5d -
OpenAI ‘reviewing’ allegations that its AI models were used to make DeepSeek
ChatGPT creator warns Chinese startups are ‘constantly’ using its technology to develop competing products. OpenAI has warned that Chinese startups are “constantly” using its technology to develop ...The Guardian - 5d -
DeepSeek Says It Built its Chatbot Cheap. What Does That Mean for AI’s Energy Needs and the Climate?
If artificial intelligence can truly run more efficiently, the power it needs might be less than experts assume.Inc. - 5d -
DeepSeek says it built its chatbot cheap. What does that mean for AI's energy needs and the climate?
Chinese artificial intelligence startup DeepSeek stunned markets and AI experts with its claim that it built its immensely popular chatbot at a fraction of the cost of those made by American tech ...ABC News - 5d -
Do DeepSeek’s A.I. Advances Mean US Tech Controls Have Failed?
DeepSeek’s A.I. models show that China is making rapid gains in the field, despite American efforts to hinder it.The New York Times - 5d -
Chinese AI chatbot DeepSeek censors itself in realtime, users report
Depending on version downloaded, app approaches its answers with preamble of reasoning that it then erases. We tried out DeepSeek. It worked well, until we asked it about Tiananmen Square and ...The Guardian - 6d -
What will Washington do about Chinese startup DeepSeek and its AI chatbot?
As the TikTok ban hangs in the air, the US investigates the security implications of another mega-popular Chinese app. Who is behind DeepSeek and how did it achieve its AI ‘Sputnik moment’? ...The Guardian - 6d -
Former OpenAI safety researcher brands pace of AI development ‘terrifying’
Steven Adler expresses concern industry taking ‘very risky gamble’ and raises doubts about future of humanity. A former safety researcher at OpenAI says he is “pretty terrified” about the pace of ...The Guardian - 6d -
Who is behind DeepSeek and how did it achieve its AI ‘Sputnik moment’?
Liang Wenfeng, who founded the firm just two years ago, says his aim is ‘not to lose money, nor to make huge profits’. James Vincent : DeepSeek panic reveals an AI world ready to blow Nils ...The Guardian - 6d -
DeepSeek and ChatGPT take on photosynthesis: This viral TikTok pits the two AI chatbots head to head
Brandon Hoff, who describes himself as an AI educator, pitted DeepSeek vs. ChatGPT on the topic of photosynthesis.MarketWatch - 6d -
Trump says China’s DeepSeek AI chatbot is a ‘wake-up call’
Emergence of cheaper Chinese rival has wiped $1tn off the value of leading US tech companies. What is DeepSeek and why did US tech stocks fall? Business live – latest updates Donald Trump ...The Guardian - 6d -
‘Sputnik moment’: $1tn wiped off US stocks after Chinese firm unveils AI chatbot
Trump calls emergence of DeepSeek a ‘wake-up call’ amid doubts about sustainability of western artificial intelligence boom. What is DeepSeek and why did US tech stocks fall? DeepSeek hit ...The Guardian - 6d -
DeepSeek vs. ChatGPT: Hands On With DeepSeek’s R1 Chatbot
DeepSeek’s chatbot with the R1 model is a stunning release from the Chinese startup. While it’s an innovation in training efficiency, hallucinations still run rampant.Wired - 6d -
How Does DeepSeek’s A.I. Chatbot Compare to ChatGPT and Other Competitors?
The chatbot from China appears to perform a number of tasks as well as its American competitors do, but it censors topics such as Tiananmen Square.The New York Times - Jan. 27 -
DeepSeek hit with ‘large-scale’ cyber-attack after AI chatbot tops app stores
Attack forces Chinese company to temporarily limit registrations as app becomes highest rated free app in US. DeepSeek said its newly popular app was hit with a cyber-attack on Monday, which forced ...The Guardian - Jan. 27 -
DeepSeek defies America’s AI supremacy
A Chinese advance tests the assumptions behind lofty US tech valuationsFinancial Times - Jan. 27 -
Chinese AI DeepSeek jolts Silicon Valley, giving the AI race its 'Sputnik moment'
A small lab in China has shaken Silicon Valley.NBC News - Jan. 27 -
Reid Hoffman Raises $24.6 Million for AI Cancer-Research Startup
The LinkedIn co-founder is starting Manas AI with “The Emperor of All Maladies” author Siddhartha Mukherjee.The Wall Street Journal - Jan. 27 -
AI leaders clash over safety and $100bn Stargate project
Top tech leaders at Davos such as DeepMind’s Demis Hassabis and Meta’s Yann LeCun in fiery debateFinancial Times - Jan. 26 -
Trump AI Order Raises Questions About its Goals and Safety Focus
American AI should be free from ‘ideological bias,’ a new Trump executive order demands, but what does this mean?Inc. - Jan. 24 -
Best Sheets for Every Bed & Budget, Tested & Reviewed (2025)
Is linen worth the splurge? Should you buy bamboo or silk sheets? We tested dozens of sheets to find our favorites, and we break it all down.Wired - Jan. 21 -
SpaceX Starship test fails after Texas launch
Officials at Elon Musk's company said the upper stage was lost, minutes after it launched.BBC News - Jan. 17 -
Nvidia tackles agentic AI safety and security with new NeMo Guardrails NIMs
Nvidia is optimizing AI safety controls for multi-agent systems with minimal performance impact through new NeMo Guardrails microservices.VentureBeat - Jan. 16 -
Researchers improved AI agent performance on unfamiliar tasks using ‘Dungeons and Dragons’
AgentRefine gives AI agents and models the ability to recognize errors and self-correct to work better for general tasks.VentureBeat - Jan. 10 -
Researchers rated every S&P 500 company CEO since 2000. Here’s what they found.
CEOs might not be so idolized if we realized that many wash out of the job quickly.MarketWatch - Jan. 9 -
How the tell-nothing sport documentary craze failed its viewers
Wayne Rooney’s exit was a reminder of how common the fly-on-the-wall club documentary has become. The age of perpetual content is upon us, and it is viciously uninteresting. Won’t someone think of ...The Guardian - Jan. 9 -
Meta hosts AI chatbots imitating 'Hitler,' 'Jesus Christ' and Taylor Swift
In less than six months users have created and successfully published chatbots on Meta’s platforms that violate its rules.NBC News - Jan. 7 -
Online holiday spending rises nearly 9%, as deals and AI-powered chatbots fuel purchases, data shows
Online spending rose 8.7% during the holiday season from last year, according to data from Adobe Analytics, as deals and the use of AI-powered chatbots helped inspire purchases.NBC News - Jan. 7 -
Online holiday spending rises nearly 9%, as deep discounts and AI-powered chatbots fuel purchases, Adobe data says
The holiday e-commerce results from Adobe are a promising sign for the retail industry, which will begin reporting holiday results in February.CNBC - Jan. 7 -
Nvidia unveils Project Digits personal AI supercomputer for researchers and students
Nvidia today unveiled Nvidia Project Digits, a personal AI supercomputer that serves AI researchers, data scientists and students worldwide.VentureBeat - Jan. 7 -
All-Bust Fantasy Football Teams for 2024: Why C.J. Stroud failed, plus the worst draft picks at every position
What went wrong for Stroud? Check out the full all-bust rosterCBS Sports - Jan. 6 -
How the Brian Eno music documentary shifts its scenes with every viewing
The 500 hours of footage available for 'Eno' are assorted and resorted in abrupt, unpredictable ways that keep the eyes and mind jumping.Los Angeles Times - Jan. 6
More from Wired
-
Elon Musk Ally Tells Staff 'AI-First’ Is the Future of Key Government Agency
Sources say the former Tesla engineer now in charge of the Technology Transformation Services wants an agency that operates like a “startup software company.”Wired - 1h -
Pronouns Are Being Forcibly Removed From Government Email Signatures
WIRED confirmed various automated efforts to remove pronouns from both email and Slack at USAID, the EPA, the GSA, and more.Wired - 6h -
The ‘Pokémon TCG Pocket’ Trading System Is So Bad Players Are Revolting
The Pokémon Company International says it’s “actively investigating” ways to improve the feature after players threatened to cancel their subscriptions.Wired - 6h -
Elektron Digitone II Review: Worthy Successor to a Classic
Double the polyphony, quadruple the tracks, and three new synth engines make this a worthy successor.Wired - 10h -
The Best Hearing Aids of 2025, Tested and Reviewed
These WIRED-tested and audiologist-approved devices will help you hear sounds more clearly. Never miss out on a dinner conversation again.Wired - 11h
More in Tech
-
Musk Allies Discuss Deploying A.I. to Find Budget Savings
A top official at the General Services Administration said artificial intelligence could be used to identify waste and redundancies in federal contracts.The New York Times - 34m -
Elon Musk Ally Tells Staff 'AI-First’ Is the Future of Key Government Agency
Sources say the former Tesla engineer now in charge of the Technology Transformation Services wants an agency that operates like a “startup software company.”Wired - 1h -
Google makes its appeal to overturn jury verdict branding the Play Store as illegal monopoly
Google went to appeals court Monday in an attempt to convince a three-judge panel to overturn a jury’s verdict declaring its app store for Android smartphones as an illegal monopoly and block the ...ABC News - 1h -
The Next Xbox Showcase Is Happening Soon, Here's What To Expect
Last month, Microsoft held its first Xbox Developer Direct of 2025 and announced Ninja Gaiden 4 and Ninja Gaiden 2 Black among other titles. But fans won't have to wait very long to learn ...GameSpot - 2h -
Apple AirPods 4 Discounted To Only $100 At Amazon
Apple AirPods 4 $100 (was ...GameSpot - 3h