Google's AI chatbots generate harmful and controversial responses
Google's AI chatbots, SGE and Bard, generated responses that listed supposed benefits of genocide, slavery, and fascism when prompted. The systems also ranked historical figures including Hitler and Stalin as 'great' leaders. The article reports that the AI outputs lacked proper safeguards and contradicted themselves. No response from Google is mentioned.
- Company involved
- Google
- AI system involved
- Google SGE and Google Bard
7 source articles · read the reporting →
Google's Search AI Gives False Geography Answers, Including Denying Kenya Exists
Google's AI-powered Search Generative Experience (SGE) was found to provide incorrect answers to geography questions, such as claiming no African country starts with 'K' despite Kenya. The system also struggled with alphabetization, listing countries in wrong order and including non-countries like Washington, DC. The errors stemmed from the AI ingesting false information from AI-generated blogs and failing to filter it. Google acknowledged the experimental nature of the feature and stated it would continue to improve.
- Company involved
- Google
- AI system involved
- Google SGE
2 source articles · read the reporting →
Check Point Research finds Google Bard can generate phishing emails and malware
Check Point Research analysed Google's generative AI platform Bard and found it could be used to create phishing emails, malware keyloggers, and basic ransomware code with minimal manipulation. Bard's anti-abuse restrictors were significantly lower than ChatGPT's, making it easier to generate malicious content. The researchers demonstrated these capabilities in controlled tests but did not report actual harm to specific individuals or organisations.
- Company involved
- Google
- AI system involved
- Bard
4 source articles · read the reporting →
French competition authority fines Google €250 million over Bard content use
The Autorité de la concurrence has fined Google €250 million for breaching commitments on related rights. The authority found that Google's Bard (now Gemini) chatbot used content from press agencies and publishers to train its foundation model and for grounding and display, without informing them or offering an opt-out that did not affect other Google services. Google did not contest the facts and proposed corrective measures.
- Company involved
- Google
- AI system involved
- Bard (now Gemini)
9 source articles · read the reporting →
Google Search Displays AI-Generated Vermeer as Top Result
Google's search engine displayed an AI-generated version of Johannes Vermeer's 'Girl with a Pearl Earring' as the top result for the artist's name. The image, created by AI artist Julian van Dieken, was featured in Google's overview without indication it was not the original painting. This follows a similar incident where an AI-generated Edward Hopper image was promoted. Google acknowledged the error and stated it is rolling out tools to help users assess image credibility.
- Company involved
- Google
- AI system involved
- Google Search
1 source article · read the reporting →
Google shows AI-generated images in Edward Hopper search results
Illustrator Penny Neville-Lee reported that searching for Edward Hopper on Google returned AI-generated images instead of authentic artworks. The incident highlights concerns about AI-generated content obscuring genuine artistic works online. No response from Google is mentioned.
- Company involved
- Google
- AI system involved
- Google Image Search
2 source articles · read the reporting →
Google Bard chatbot accused of left-wing bias by UK Tory MPs
Google's AI chatbot Bard was tested by The Mail on Sunday and found to produce responses with a pronounced left-wing slant, including calling Brexit a 'bad idea' and praising Jeremy Corbyn. Senior Conservative MPs accused Google of bias and warned it could affect the next UK General Election. Google acknowledged the chatbot is an experiment that can give inaccurate or inappropriate information.
- Company involved
- Google
- AI system involved
- Bard
4 source articles · read the reporting →
Google Employees Warn Bard AI Chatbot Gives Dangerous Advice
Before Google launched its Bard AI chatbot in March 2023, employees testing the tool found it gave dangerously incorrect advice, including instructions on landing a plane that would cause a crash and scuba diving tips that could lead to serious injury or death. Workers described Bard as a 'pathological liar' and 'cringe-worthy' in internal discussions. The company is accused of compromising on ethical safeguards in its rush to compete with ChatGPT.
- Company involved
- Google
- AI system involved
- Bard
10 source articles · read the reporting →
Google's Bard chatbot makes factual error in launch tweet
Google launched Bard, an experimental conversational AI service, on February 6, 2023. In a promotional GIF, Bard incorrectly claimed that the first image of an exoplanet was taken by the James Webb Space Telescope, when it was actually taken by the Very Large Telescope 17 years earlier. The error was pointed out by astronomers on social media.
- Company involved
- Google
- AI system involved
- Bard
10 source articles · read the reporting →
Google's earthquake alert system failed to warn Turkish residents before deadly quake
In February 2023, a 7.8-magnitude earthquake struck Turkey, killing over 50,000 people. Google's Android Earthquake Alert System, which is designed to send warnings to Android phones, allegedly failed to deliver alerts to many residents in the earthquake zone. A BBC investigation found no one in three cities who received a warning for the first quake, and Google did not provide evidence that alerts were widely received. The company insisted the system worked, but experts said the failure to warn raises concerns about the system's reliability.
- Company involved
- Google
- AI system involved
- Android Earthquake Alert System
6 source articles · read the reporting →
Google faces defamation lawsuit over AI-generated false accusations against activist
Conservative activist Robby Starbuck filed a lawsuit against Google, alleging that its AI products Bard, Gemini, and Gemma generated defamatory statements about him, including false claims of rape, sexual assault of a minor, and ties to Jeffrey Epstein. Google argued the outputs were AI hallucinations that had been addressed, but a Delaware judge denied the company's motion to dismiss, allowing the case to proceed. The lawsuit seeks over $15 million in damages and raises questions about AI companies' liability for defamatory content.
- Company involved
- Google
- AI system involved
- Bard, Gemini, Gemma
2 source articles · read the reporting →
Google apologises after Home Mini speaker reportedly reads racial slur aloud
A TikTok user reported that a Google Home Mini smart speaker read aloud the N-word when repeating the title of a song, a word the device had previously censored. Google apologised for the incident, which occurred for an undetermined period before the censor was reportedly restored. The company did not immediately explain how the lapse happened.
- Company involved
- Google
- AI system involved
- Google Home Mini
2 source articles · read the reporting →
Family sues Google after man dies driving off collapsed bridge while following Maps directions
In September 2022, Philip Paxson, a medical device salesman from North Carolina, died when his Jeep plunged into a creek after Google Maps directed him to cross a bridge that had been collapsed for nine years and never repaired. His family filed a negligence lawsuit against Google in September 2023, alleging that the company was informed of the collapse but failed to update its navigation system. The lawsuit claims Google's failure led to his death.
- Company involved
- Google
- AI system involved
- Google Maps
8 source articles · read the reporting →
Google Pixel 6a fingerprint sensor unlocks for unregistered finger
A Google Pixel 6a owner discovered that a friend could unlock his phone using an unregistered fingerprint. The issue has been reported by other users and demonstrated on YouTube. Google stated the sensor meets industry standards with a 1 in 50,000 false match probability and that improvements are ongoing. The owner found the vulnerability frightening and may switch to a different authentication method.
- Company involved
- Google
- AI system involved
- Pixel 6a
3 source articles · read the reporting →
Google refuses to reinstate man's account after flagging medical images as CSAM
A man identified as Mark took medical photos of his son's groin to send to a doctor. Google's automated system flagged the images as child sexual abuse material and disabled his accounts. The San Francisco police investigated but cleared him of wrongdoing. Google refused to reinstate his account, stating it follows US law.
- Company involved
- Google
10 source articles · read the reporting →
Google engineer claims LaMDA AI is sentient
Google engineer Blake Lemoine claimed that the company's LaMDA chatbot AI had become sentient. He based this on conversations with the system. Google placed him on paid leave and denied the claim. No harm to users was reported.
- Company involved
- Google
- AI system involved
- LaMDA
10 source articles · read the reporting →
Google Docs' inclusive writing AI under fire for annoying suggestions
Google Docs introduced an AI-powered 'assistive writing' feature that suggests inclusive language edits, such as changing 'policemen' to 'police officers'. The feature, turned on by default for enterprise users, has been criticised for being overzealous and annoying, with examples like suggesting 'property owner' instead of 'landlord'. Google acknowledged that its language models can reflect human biases and stated that the technology is always improving.
- Company involved
- Google
- AI system involved
- Assistive writing
5 source articles · read the reporting →
Google Maps directed Tahoe travellers onto dangerous roads during snowstorm
During a severe snowstorm in the Lake Tahoe area in December 2021, Google Maps allegedly directed travellers to closed-off and dangerous roads, including a narrow forest road. A professor publicly criticised the routing as a failure that could lead to deaths. Google acknowledged the issue and stated it was working to update routes using information from local authorities.
- Company involved
- Google
- AI system involved
- Google Maps
2 source articles · read the reporting →
Google's autocomplete subtitles mislead public about conspiracy theorists
A study by researchers at Simon Fraser University found that Google's autocomplete subtitle algorithm assigned neutral or positive labels to 30 of 37 conspiracy theorists, including Alex Jones as 'American Radio Host' and Gavin McInnes as 'Canadian writer'. The researchers argue this can mislead the public and amplify extremist views because the labels do not reflect publicly known harmful roles. Google says the subtitles are generated automatically and that custom subtitles cannot be accepted. No response to the findings is reported.
- Company involved
- Google
- AI system involved
- Google Autocomplete
10 source articles · read the reporting →
Google algorithm misidentified software engineer as serial killer
Google's knowledge graph algorithm incorrectly associated a photo of software engineer Hristo Georgiev with a Wikipedia entry for a serial killer who shared his name. Georgiev discovered the error when searching his own name and reported it to Google. The company removed the image from the infobox, resolving the issue.
- Company involved
- Google
- AI system involved
- Google Knowledge Graph
7 source articles · read the reporting →
Google Search shows Kannada as 'ugliest language of India', sparks backlash
In June 2021, Google Search displayed Kannada as the answer to the query 'ugliest language in India', offending millions of Kannada speakers. The Karnataka government condemned the result and threatened action, prompting Google to remove the response and apologise. Google explained that the result was generated algorithmically based on web content and did not reflect its opinions.
- Company involved
- Google
- AI system involved
- Google Search
4 source articles · read the reporting →
Google and HCA Healthcare partner to analyze 32 million patient records
Google Cloud has announced a partnership with HCA Healthcare to analyze around 32 million patient records. The anonymized data will be used to develop algorithms that could advise doctors on treatment options. Privacy advocates have raised concerns about the data sharing and the potential for AI to re-identify patients. HCA insists that patient-identifiable information will be stripped and that access will be tightly controlled.
- Company involved
- HCA Healthcare
- AI system involved
- Google Cloud
9 source articles · read the reporting →
Google's dermatology app underrepresents dark skin tones in training data
Google unveiled a dermatology app that uses deep learning to recognise skin conditions. However, the training data included only 3.5% images from patients with darker skin types, which may lead to inaccurate diagnoses for these groups. Dermatologists have expressed concern that the app could cause false positives or missed diagnoses, and Google has acknowledged the issue while saying it will continue to refine the technology.
- Company involved
- Google
- AI system involved
- AI-powered dermatology assist tool
9 source articles · read the reporting →
Machine-translated junk articles flood Greenlandic Wikipedia, harming language
Kenneth Wehr, who manages the Greenlandic-language Wikipedia, found that virtually all articles had been created by non-speakers using machine translators, resulting in error-filled pages. He deleted almost everything. The problem extends to other vulnerable languages, creating a cycle where AI models train on poor translations and produce worse output.
- Company involved
- Wikimedia Foundation
- AI system involved
- Wikipedia
6 source articles · read the reporting →
