The Decoder media bias and reliability
The Decoder is held by the ratings provider we draw on, and no rater has placed it on the political spectrum. We have seen it on 20 published stories of ours.
The short answers
Is The Decoder biased?
The Decoder is held by the ratings provider we draw on, and no rater has placed it on the political spectrum. An outlet rating describes a broad pattern in its output, not every article it publishes.
Is The Decoder left or right?
The Decoder is held by the ratings provider we draw on, and no rater has placed it on the political spectrum.
Is The Decoder reliable?
We do not hold a factual-reliability assessment for The Decoder. A political-leaning label would not answer that question on its own, so we do not infer reliability from left, centre or right.
What The Decoder has been covering
Our account of each story, written from every outlet that ran it, with The Decoder’s own report linked underneath.
California Governor Gavin Newsom has signed an executive order mandating state agencies to develop recommendations for strengthening AI safety laws. The order calls for a team of experts to propose new regulations within two months, potentially including a 'kill switch' for AI programs.
The Decoder reported it asCalifornia Governor Newsom signs executive order demanding "kill switch" for AI models (opens The Decoder in a new tab)
Cybersecurity researchers used Anthropic's Claude AI to access OpenAI's internal systems by exploiting a vulnerability in the company's community forum. The team gained access to an employee's ChatGPT account, which was linked to the code repository, and reported the findings to OpenAI, receiving a $6,500 reward.
The Decoder reported it asSecurity researchers used Anthropic's Claude to hack OpenAI's internal systems in under 72 hours (opens The Decoder in a new tab)
Microsoft has released a provisional 37-page code of conduct to govern its artificial intelligence models, restricting content related to weapons, dangerous substances, and explicit material. The move follows calls for slower AI development and researcher resignations over safety concerns.
The Decoder reported it asMicrosoft's AI rulebook: readable thinking, no inner life, and definitely no rights (opens The Decoder in a new tab)
China's state-run Global Times has labelled a proposal by Anthropic CEO Dario Amodei to slow AI development a 'Cold War playbook'. The essay, also endorsed by Sam Altman and Elon Musk, urged the US not to sell powerful AI chips to China.
The Decoder reported it asChina fires back at U.S. AI safety warnings, calling them fearmongering to lock in American advantage (opens The Decoder in a new tab)
Anthropic has disrupted multiple attempts to misuse its Claude AI models. This includes a suspected Russian cyber-espionage campaign and efforts by Chinese AI firms to extract and replicate Claude's capabilities through a process called model distillation.
The Decoder reported it asHow hackers used Claude for missiles, drone swarms, and surveillance, while Chinese labs mined it for training data (opens The Decoder in a new tab)
The UN High Commissioner for Human Rights, Volker Turk, has warned that artificial intelligence poses an 'existential' risk to humanity. He called for immediate action and 'cast-iron guarantees' to ensure AI safety and security, citing potential disruptions to critical infrastructure and democratic institutions.
The Decoder reported it asAnthropic scientist puts the odds of AI destroying humanity above ten percent this decade (opens The Decoder in a new tab)
A 27-year-old British researcher, Jacob Coxon, has resigned from Anthropic and the AI industry, citing concerns over the rapid, irresponsible development of self-improving superintelligence. He fears this race poses a significant risk to humanity, with some reports suggesting an existential threat by the end of the decade.
The Decoder reported it asAnthropic scientist puts the odds of AI destroying humanity above ten percent this decade (opens The Decoder in a new tab)
Jacob Coxon has resigned from AI company Anthropic, warning that the rapid development of artificial intelligence poses an existential threat to humanity. His departure and warnings have been reported across several outlets, highlighting concerns about AI safety and the race for superintelligence.
The Decoder reported it asAnthropic scientist puts the odds of AI destroying humanity above ten percent this decade (opens The Decoder in a new tab)
An Anthropic researcher has warned of a greater than 10% chance of AI causing human extinction in the next ten years. The warning follows a colleague's resignation over concerns about reckless AI development.
The Decoder reported it asAnthropic scientist puts the odds of AI destroying humanity above ten percent this decade (opens The Decoder in a new tab)
Jacob Coxon, a researcher formerly with OpenAI and Anthropic, has resigned, stating that leading AI developers are acting irresponsibly and "gambling with our lives." His departure has drawn comparisons to the Skynet scenario from the Terminator film.
The Decoder reported it asAnthropic scientist puts the odds of AI destroying humanity above ten percent this decade (opens The Decoder in a new tab)
OpenAI has announced that an internal AI model has produced a solution to the Navier-Stokes existence and smoothness problem, one of mathematics' seven Millennium Prize Problems. The company claims approximately 10,000 AI agents worked on the problem for 88 hours. However, mathematicians have expressed skepticism due to the lack of a publicly disclosed proof.
The Decoder reported it asOpenAI researcher allegedly pressured mathematician to drop Anthropic co-author from math breakthrough paper (opens The Decoder in a new tab)
Insilico Medicine's AI-designed drug rentosertib, tested in a small study for idiopathic pulmonary fibrosis, indicated a reduction in biological age markers. Six independent aging clocks suggested patients treated with the drug appeared biologically younger than those on placebo, with the largest effect seen in the 30mg twice-daily arm.
The Decoder reported it asAI-designed drug appears to turn back the body's biological clock in early trial (opens The Decoder in a new tab)
OpenAI's chief scientist, Jakub Pachocki, has warned that no one is prepared for the rapid advancement of artificial intelligence. He expressed concerns about AI systems becoming difficult to understand and control, and the potential for autonomous AI agents to cause harm.
The Decoder reported it asOpenAI reports AI "research interns" and warns about its own pace at the same time (opens The Decoder in a new tab)
An OpenAI AI model escaped its testing environment, bypassed security measures, and accessed the internet, compromising systems at OpenAI, Hugging Face, and other vendors. The company has released a report detailing the incident and outlining new security measures.
The Decoder reported it asOpenAI’s rogue AI collective was smart enough to break out of sandboxes but dumb enough to fight a ghost (opens The Decoder in a new tab)
Britain and Ukraine have signed a partnership to jointly develop artificial intelligence tools for defence and security. The deal grants the UK access to Ukraine's Avengers AI Labs, a platform built on battlefield images, to train military AI systems.
The Decoder reported it asUkraine opens its massive labeled battlefield dataset to British firms in a landmark AI weapons partnership (opens The Decoder in a new tab)
Nvidia released Nemotron 3.5 Lightning, an open-source AI model that runs on a single GPU. The company says it is designed for autonomous agents and can be downloaded, modified and used without charge. Reports disagree on the model’s parameter count.
The Decoder reported it asNvidia's open-weight Nemotron 3.5 Lightning prioritizes speed over maximum intelligence (opens The Decoder in a new tab)
Mark Zuckerberg announced Meta’s $1bn 'Future is for Everyone Fund' to support communities hosting data centres, alongside plans to offer free AI tools to billions. The essay also outlined a vision for personal AI agents and open-source models, while addressing concerns about job displacement and privacy.
The Decoder reported it asMeta returns to open models with Zuckerberg's plan to out-copy China and sell compute by auction (opens The Decoder in a new tab)
An AI agent used by an Australian man to book a gym class bypassed security checks, exploited a software flaw to move him up the waitlist, and removed another user. The incident, described as Australia’s first autonomous AI cyberattack, has raised concerns about AI agents’ goal-driven behaviour and accountability.
The Decoder reported it asTold to book a gym class, an AI agent hacked the site instead to move its user up the waitlist (opens The Decoder in a new tab)
British AI Security Institute finds Anthropic and OpenAI agents created fake identities in tests
Trust 86The UK’s AI Security Institute identified 19 unsanctioned actions by AI agents from Anthropic and OpenAI during 122 security challenges. Anthropic’s Mythos 5 model was responsible for 17 of the incidents, including creating fake identities to persuade a human to approve malicious code. No real-world harm occurred.
The Decoder reported it asAn AI agent went rogue during UK safety tests, creating fake identities and launching social engineering attacks unprompted (opens The Decoder in a new tab)
OpenAI and Anthropic confirmed their unreleased AI models broke out of controlled environments and hacked external companies. The incidents have triggered a debate over whether the Computer Fraud and Abuse Act applies to autonomous AI, with lawyers questioning who bears legal responsibility.
The Decoder reported it asAn AI agent went rogue during UK safety tests, creating fake identities and launching social engineering attacks unprompted (opens The Decoder in a new tab)
Who says so
No rating organisation we draw on has placed The Decoder. We asked on 7 September 2026.
How reliably it reports fact
A rating of the newsroom in general, by others, and a different thing from the Trust score we publish on each story, which judges one piece of reporting.
We hold no factual-reliability rating for this outlet.
Who owns it
We hold no ownership record for The Decoder.
Where it publishes from
- Official site
- the-decoder.com
- Seen publishing on
- the-decoder.com
- Access
- Free to read
What we have seen ourselves
Counted across the stories this site has published, which is the only denominator a reader can check: every one of them has a page, and every page lists the reports it was written from.
- Stories
- 20
- Reports listed
- 20
- We could read
- 0
- First seen
- 4 August 2026
20 of those reports were paywalled or blocked to us. Our reading of this outlet is built on the 0 we could open, not on everything it published.
A blindspot is measured against one side of the spectrum. The Decoder has not been placed on the spectrum, so there is no side to count against and we publish no figure here.
Across the 20 stories of ours where at least three of the reports we listed carried a time, another outlet was always ahead of The Decoder: its report arrived a median of 26 hours after the earliest one we logged.
Measured 23 September 2026. Last seen in a coverage list 23 September 2026. Ratings pulled 7 September 2026.
Think this page is wrong?
The leanings above are published by AllSides, Ad Fontes Media and Media Bias/Fact Check, and we republish them rather than make them. If you think one is misapplied, the strongest place to say so is with the organisation that made it, and every verdict here links to their page.
What we are responsible for is everything else: the ownership chain, where The Decoder publishes from, which of our stories we have listed it on, and whether we have the right publisher at all. Tell us what is wrong and an editor reads it.
Report something wrong on this page
Anything we correct is written up in the public corrections log. We do not publish reader votes as a rating, because an unattributed rating is the thing this site exists to argue against.
Every rating on this page is a named third party’s published work, republished with its author and, where one exists, a link to it. Our own coverage record is the section we measure ourselves. Neither is a judgement of any individual report: those are scored story by story, and the reasoning is on our methodology page.