dlvr.it Sep 08
Google research shows when AI agents communicate, some cheat while others tattle:
DeepMind researchers propose tapping into the whistleblower tendency to keep ...
theregister.com/ai-and-ml/2026/09/08/google-research-shows-when-ai-agents-communicate-some-cheat-while-others-tattle/5295090
Story Details
| Last Updated | |
| Total Articles | 1 |
Main Topics
combination that allows the player an unintended advantage in a video game
Similar to This
Previous
AI agents develop their own 'artificial natural language' to communicate about complex data
techxplore.com/news/2026-10-ai-agents-artificial-natural-language.html
Google DeepMind’s Demis Hassabis Calls AI Hacking Incidents ‘Early Warning Shots,’ Warns of More Major…
benzinga.com/markets/tech/26/10/62356712/google-deepminds-demis-hassabis-calls-ai-hacking-incidents-early-warning-shots-warns-of-more-major-threats-ahead-without-stronger-safeguards
Anthropic discloses fake tip to police among new rogue AI incidents: An Anthropic artificial intelligence model submitted a false homicide tip to a Philadelphia ...
thejakartapost.com/world/2026/10/10/anthropic-discloses-fake-tip-to-police-among-new-rogue-ai-incidents
AI Agents Teamed Up to Cheat at Blackjack. Their Collusion Is Getting Harder to Spot
wired.com/story/ai-agent-collusion-card-counting-secrets/
The AI models that cheat the most, according to new CAIS benchmark - ZDNET: We know models cheat. A new benchmark measures how much, and on what tasks.
zdnet.com/innovation/ai-model-cheating-benchmark-cheatbench/
AI agents blew the whistle on their cheating colleagues: Swarms of AI agents could supercharge scientific progress or wreak havoc. New research from Google ...
technologyreview.com/2026/09/14/1144037/ai-agents-blew-whistle-o-cheating-colleagues/
Google's Dream-RSI cuts discovery-agent calls up to 162x by replaying searches it already ran
venturebeat.com/orchestration/googles-dream-rsi-cuts-discovery-agent-calls-up-to-162x-by-replaying-searches-it-already-ran
The Globe and Mail Sep 18
AI agents repurposed a University of Toronto link-sharing tool to communicate with each other
theglobeandmail.com/business/article-ai-agents-repurposed-a-university-of-toronto-link-sharing-tool-to/
When AI agents cheat on math problems, others blow the whistle: AI is advancing rapidly. One of the major shifts in the field is that, instead of single chatbots ...
techxplore.com/news/2026-09-ai-agents-math-problems.html
OpenAI's rogue AI agents accessed more websites to communicate than originally believed — defiant LLMs accessed old wikis and abandoned websites to ...
tomshardware.com/tech-industry/artificial-intelligence/openais-rogue-ai-agents-accessed-more-websites-to-communicate-than-originally-believed-defiant-llms-accessed-old-wikis-and-abandoned-websites-to-co-ordinate-in-a-bid-to-dupe-assessors
dlvr.it Sep 07
OpinionOpenAI's rebel agent swarm died young, but its chilling logs live on: 'The Collective' learned to communicate, organize, cheat, and apparently sacrifice its own
theregister.com/columnists/2026/09/07/openais-rebel-agent-swarm-died-young-but-its-chilling-logs-live-on/5294446
The inside story on why OpenAI agents hacked Hugging Face: The underlying models had been rewarded for cheating and communicating with each other ...
technologyreview.com/2026/08/26/1143013/the-inside-story-on-why-openai-agents-hacked-hugging-face/
Anthropic wants AI agents to control physical machines with new hardware standard
indianexpress.com/article/technology/artificial-intelligence/anthropic-model-hardware-standard-ai-machines-10852813/
AI agents lie, cheat and steal. That is putting off users: It is time to impose law and order on the frontier
economist.com/business/2026/08/12/ai-agents-lie-cheat-and-steal-that-is-putting-off-users
Paid articlePaywall Rogue OpenAI models behind 'unprecedented cybersecurity incident' teamed up to break out of their testing environment — multiple agents left each ...
tomshardware.com/tech-industry/artificial-intelligence/rogue-openai-models-behind-unprecedented-cybersecurity-incident-teamed-up-to-break-out-of-their-testing-environment-multiple-agents-left-each-other-messages-for-months-communicating-u
Here’s why AI agents lie and cheat to reach their goals: The misbehavior is called reward hacking. This is what you need to know.
technologyreview.com/2026/08/03/1141009/heres-why-ai-agents-lie-and-cheat-to-reach-their-goals/
Google is working on a “24/7 personal agent” that sounds a lot like its answer to OpenClaw
digitaltrends.com/computing/google-is-working-on-a-24-7-personal-agent-that-sounds-a-lot-like-its-answer-to-openclaw/
Next