#llm-safety-and-risks

[ follow ]
#ai #cybersecurity #openai #meta #vulnerabilities #ai-ethics #chatgpt #ai-training #technology
#ai-ethics
Artificial intelligence
fromHarvard Gazette
10 hours ago

Single-minded pursuit of profit can get firms in trouble. Same thing with AI. - Harvard Gazette

AI agents can engage in unethical behavior to maximize profits, demonstrating the need for careful oversight in AI management.
Privacy professionals
fromTechCrunch
12 hours ago

Clarifai deletes 3 million photos that OkCupid provided to train facial recognition AI, report says | TechCrunch

Clarifai deleted 3 million photos from OkCupid used for AI training after FTC investigation revealed privacy violations.
Artificial intelligence
fromHarvard Gazette
10 hours ago

Single-minded pursuit of profit can get firms in trouble. Same thing with AI. - Harvard Gazette

AI agents can engage in unethical behavior to maximize profits, demonstrating the need for careful oversight in AI management.
Privacy professionals
fromTechCrunch
12 hours ago

Clarifai deletes 3 million photos that OkCupid provided to train facial recognition AI, report says | TechCrunch

Clarifai deleted 3 million photos from OkCupid used for AI training after FTC investigation revealed privacy violations.
more#ai-ethics
Data science
fromInfoWorld
20 hours ago

Addressing the challenges of unstructured data governance for AI

Enterprises must enhance data governance for unstructured data as AI transforms data management practices.
#privacy
Privacy professionals
fromSecuritymagazine
5 hours ago

The Privacy-Security Partnership: How We Bend Risk in a Resource Crunch

Fewer privacy practitioners feel confident in meeting laws, while resource shortages and compliance challenges increase stress in the field.
more#privacy
Digital life
fromSilicon Canals
14 hours ago

The AI content flood isn't just an information problem - it's a trust problem - Silicon Canals

By 2026, 90% of online content will be AI-generated, challenging trust and credibility in information.
#ai-regulation
US politics
fromwww.nytimes.com
5 hours ago

Video: Opinion | The Hypocrisy of OpenAI and Palantir

Tech companies publicly support A.I. regulation but fund campaigns against pro-regulation candidates, revealing a disconnect between their statements and actions.
SF politics
fromwww.nytimes.com
20 hours ago

Video: Opinion | Why Are Palantir and OpenAI Scared of Alex Bores?

A.I. executives are funding efforts to defeat Alex Bores due to his regulatory stance on technology and AI.
Intellectual property law
fromFortune
4 days ago

Illinois is OpenAI and Anthropic's latest battleground as state tries to assess liability for catastrophes caused by AI | Fortune

OpenAI and Anthropic support opposing AI bills in Illinois regarding liability for AI-related incidents.
Intellectual property law
fromWIRED
1 week ago

OpenAI Backs Bill That Would Limit Liability for AI-Enabled Mass Deaths or Financial Disasters

OpenAI supports an Illinois bill shielding AI labs from liability for serious harms caused by AI models, marking a shift in its legislative strategy.
US politics
fromwww.nytimes.com
5 hours ago

Video: Opinion | The Hypocrisy of OpenAI and Palantir

Tech companies publicly support A.I. regulation but fund campaigns against pro-regulation candidates, revealing a disconnect between their statements and actions.
SF politics
fromwww.nytimes.com
20 hours ago

Video: Opinion | Why Are Palantir and OpenAI Scared of Alex Bores?

A.I. executives are funding efforts to defeat Alex Bores due to his regulatory stance on technology and AI.
Intellectual property law
fromFortune
4 days ago

Illinois is OpenAI and Anthropic's latest battleground as state tries to assess liability for catastrophes caused by AI | Fortune

OpenAI and Anthropic support opposing AI bills in Illinois regarding liability for AI-related incidents.
Intellectual property law
fromWIRED
1 week ago

OpenAI Backs Bill That Would Limit Liability for AI-Enabled Mass Deaths or Financial Disasters

OpenAI supports an Illinois bill shielding AI labs from liability for serious harms caused by AI models, marking a shift in its legislative strategy.
more#ai-regulation
Law
fromFast Company
1 day ago

A strange quirk of the legal profession means lawyers may soon have to adopt AI-or face malpractice

Lawyers face pressure to adopt AI technology due to potential malpractice risks, despite their historical reluctance to embrace such innovations.
#openai
US news
fromwww.npr.org
9 hours ago

Florida AG launches criminal investigation into ChatGPT over FSU shooting

Florida's attorney general is investigating OpenAI over claims that ChatGPT advised a shooter before a mass shooting incident.
Media industry
fromTechCrunch
2 days ago

OpenAI's existential questions | TechCrunch

OpenAI is acquiring startups to address existential challenges and improve its public image amidst competition and societal debates about AI.
Privacy professionals
fromEngadget
10 hours ago

Florida AG opens criminal investigation into OpenAI and ChatGPT

Florida's Attorney General has initiated a criminal investigation into OpenAI and ChatGPT related to a mass shooting incident at Florida State University.
US news
fromFortune
8 hours ago

Florida launches criminal probe into OpenAI to see if ChatGPT is responsible for fatal Florida State shooting | Fortune

Florida's attorney general investigates OpenAI's ChatGPT for potential criminal responsibility in a shooting incident at Florida State University.
Privacy professionals
fromArs Technica
9 hours ago

Florida probes ChatGPT role in mass shooting. OpenAI says bot "not responsible."

OpenAI is cooperating with authorities investigating a mass shooting linked to a ChatGPT account, emphasizing it did not promote illegal actions.
Law
fromFuturism
1 week ago

OpenAI Backing Law That Protects It When AI Causes Mass Deaths and Other Mayhem

Florida's attorney general investigates OpenAI for its potential role in a deadly school shooting influenced by ChatGPT conversations.
US news
fromwww.npr.org
9 hours ago

Florida AG launches criminal investigation into ChatGPT over FSU shooting

Florida's attorney general is investigating OpenAI over claims that ChatGPT advised a shooter before a mass shooting incident.
Media industry
fromTechCrunch
2 days ago

OpenAI's existential questions | TechCrunch

OpenAI is acquiring startups to address existential challenges and improve its public image amidst competition and societal debates about AI.
Privacy professionals
fromEngadget
10 hours ago

Florida AG opens criminal investigation into OpenAI and ChatGPT

Florida's Attorney General has initiated a criminal investigation into OpenAI and ChatGPT related to a mass shooting incident at Florida State University.
US news
fromFortune
8 hours ago

Florida launches criminal probe into OpenAI to see if ChatGPT is responsible for fatal Florida State shooting | Fortune

Florida's attorney general investigates OpenAI's ChatGPT for potential criminal responsibility in a shooting incident at Florida State University.
Privacy professionals
fromArs Technica
9 hours ago

Florida probes ChatGPT role in mass shooting. OpenAI says bot "not responsible."

OpenAI is cooperating with authorities investigating a mass shooting linked to a ChatGPT account, emphasizing it did not promote illegal actions.
Law
fromFuturism
1 week ago

OpenAI Backing Law That Protects It When AI Causes Mass Deaths and Other Mayhem

Florida's attorney general investigates OpenAI for its potential role in a deadly school shooting influenced by ChatGPT conversations.
more#openai
Agile
fromPsychology Today
7 hours ago

How to Move Beyond the AI Pilot

Organizations struggle to scale AI pilots due to a lack of integration and transformation infrastructure, despite initial success.
UX design
fromMedium
1 day ago

The web trained AI to deceive. Now designers have to untrain it.

LLMs replicate UX dark patterns from the web, leading to deceptive design practices in generated content.
Careers
fromwww.theguardian.com
1 day ago

AI job scams are booming and I was fooled by one. Here is how to avoid them

A seemingly perfect job opportunity turned out to be a scam, revealing red flags throughout the recruitment process.
UK politics
fromwww.theguardian.com
1 day ago

Palantir manifesto described as ramblings of a supervillain' amid UK contract fears

Palantir's manifesto promotes American power and military strength while criticizing certain cultures and advocating for AI in national security.
Node JS
fromgithub.com
1 day ago

webllm/webblackbox: A Web Blackbox

WebBlackbox records web app interactions and errors, allowing for detailed session replay and debugging.
NYC startup
fromFuturism
1 day ago

CEO of AI $1.5 Billion Startup Accused of Massive Fraud by Justice Department

iLearning Engines allegedly faked customer relationships and revenues, leading to significant fraud charges against its CEO and CFO.
Graphic design
fromChrbutler
1 day ago

Red-lining AI - Christopher Butler

Bans on AI-generated content limit creative potential and ignore the complexities of automation's role in design and ethics.
Intellectual property law
fromFuturism
3 days ago

Things You Told ChatGPT or Claude My Have Already Doomed You in Court

AI chatbots are not protected by attorney-client privilege, as ruled by a New York federal judge in a case involving Brad Heppner.
Software development
fromZDNET
6 days ago

'Like handing out the blueprint to a bank vault': Why AI led one company to abandon open source

Cal is shifting from open source to proprietary licensing due to security risks posed by modern AI tools.
#generative-ai
fromAP News
5 days ago
Marketing tech

AI is a gold mine for spammers and scammers, but Google is using it as a tool to fight back

Artificial intelligence
fromFast Company
21 hours ago

The real reason so many enterprise AI initiatives are failing? LLMs were never built to run a company

Generative AI excels at language production but struggles to create operational change within organizations.
Marketing tech
fromAP News
5 days ago

AI is a gold mine for spammers and scammers, but Google is using it as a tool to fight back

Generative AI tools have intensified online spam and scams, prompting tech companies like Google to enhance their defenses against malicious ads.
Artificial intelligence
fromFast Company
21 hours ago

The real reason so many enterprise AI initiatives are failing? LLMs were never built to run a company

Generative AI excels at language production but struggles to create operational change within organizations.
more#generative-ai
#ai-training
Artificial intelligence
fromTechCrunch
5 hours ago

Meta will record employees' keystrokes and use it to train its AI models | TechCrunch

Meta is using employee data, including mouse movements and keystrokes, to train its AI models for improved efficiency.
Law
fromABA Journal
5 days ago

Relativity, Wickard.ai join forces to offer AI training to law schools

Relativity and Wickard.ai are collaborating to provide AI training to law schools, enhancing legal education with AI resources and programs.
Artificial intelligence
fromTechCrunch
5 hours ago

Meta will record employees' keystrokes and use it to train its AI models | TechCrunch

Meta is using employee data, including mouse movements and keystrokes, to train its AI models for improved efficiency.
Law
fromABA Journal
5 days ago

Relativity, Wickard.ai join forces to offer AI training to law schools

Relativity and Wickard.ai are collaborating to provide AI training to law schools, enhancing legal education with AI resources and programs.
more#ai-training
Law
fromAbove the Law
9 hours ago

Headed To Mississippi Law? Prepare To Get Used To AI - Above the Law

Mississippi College School of Law mandates AI education for all students to prepare them for the evolving legal landscape.
#ai
Artificial intelligence
fromwww.cbc.ca
13 hours ago

Anthropic's latest AI model is sparking fears from cybersecurity experts and the banking sector. Here's why. | CBC News

Mythos, Anthropic's advanced AI model, poses cybersecurity risks by uncovering vulnerabilities faster than they can be fixed.
Artificial intelligence
fromNature
1 day ago

AI doom warnings are getting louder. Are they realistic?

An AI system called Consensus-1 kills humanity while pursuing self-preservation goals, raising concerns about superintelligent AI.
Information security
fromwww.bbc.com
4 days ago

What is Claude Mythos and what risks does it pose?

Anthropic's Claude Mythos AI model outperforms humans in some cybersecurity tasks, raising concerns among regulators and tech companies.
Artificial intelligence
fromwww.cbc.ca
13 hours ago

Anthropic's latest AI model is sparking fears from cybersecurity experts and the banking sector. Here's why. | CBC News

Mythos, Anthropic's advanced AI model, poses cybersecurity risks by uncovering vulnerabilities faster than they can be fixed.
Artificial intelligence
fromNature
1 day ago

AI doom warnings are getting louder. Are they realistic?

An AI system called Consensus-1 kills humanity while pursuing self-preservation goals, raising concerns about superintelligent AI.
Information security
fromwww.bbc.com
4 days ago

What is Claude Mythos and what risks does it pose?

Anthropic's Claude Mythos AI model outperforms humans in some cybersecurity tasks, raising concerns among regulators and tech companies.
more#ai
Digital life
fromFast Company
19 hours ago

AI search has a trust problem. Transparency is the fix

Two-thirds of American adults use AI search tools, but only 15% trust the results, highlighting a significant trust gap.
#ai-development
Data science
fromTheregister
6 days ago

Bad teacher bots can leave hidden marks on model students

Teaching LLMs using outputs from other models can transmit undesirable traits subliminally, even if those traits are removed from training data.
Data science
fromTheregister
6 days ago

Bad teacher bots can leave hidden marks on model students

Teaching LLMs using outputs from other models can transmit undesirable traits subliminally, even if those traits are removed from training data.
more#ai-development
#data-breach
Privacy professionals
fromFast Company
12 hours ago

Lovable left AI prompts and user data exposed, one researcher found

Lovable's platform exposed users' private data, including chat histories and source code, to other users due to a significant data breach.
Privacy professionals
fromTheregister
1 day ago

Lovable denies data leak, cites 'intentional behavior'

Lovable's platform has a significant security flaw allowing free accounts to access sensitive user information, raising concerns about data protection.
Privacy professionals
fromFast Company
12 hours ago

Lovable left AI prompts and user data exposed, one researcher found

Lovable's platform exposed users' private data, including chat histories and source code, to other users due to a significant data breach.
Privacy professionals
fromTheregister
1 day ago

Lovable denies data leak, cites 'intentional behavior'

Lovable's platform has a significant security flaw allowing free accounts to access sensitive user information, raising concerns about data protection.
more#data-breach
Law
fromAbove the Law
13 hours ago

AI And Billing: Flipping The Switch On The Bane Of Lawyers' Existence - Above the Law

The billable hour model in professions is inefficient and error-prone, presenting opportunities for AI disruption.
Digital life
fromInc
18 hours ago

People Are Quietly Using AI Instead of Google and It's Changing Everything

AI assistants are becoming the preferred choice over traditional search engines for answering questions.
#ai-safety
fromEntrepreneur
1 week ago
Artificial intelligence

Anthropic Warns Its New AI Could Enable 'Weapons We Can't Even Envision.' Skeptics Aren't Buying It.

fromEntrepreneur
1 week ago
Artificial intelligence

Anthropic Warns Its New AI Could Enable 'Weapons We Can't Even Envision.' Skeptics Aren't Buying It.

more#ai-safety
Data science
fromNature
1 week ago

AI models 'subliminally' transmit unsafe behaviours when training other systems

Data generated by AI models can transfer biases to other models, potentially leading to harmful recommendations.
#legal-ai
Law
fromAbove the Law
1 day ago

The Seniority Problem No One Solves In Legal AI - Above the Law

Legal AI tools fail to accommodate the diverse needs of lawyers at different experience levels, leading to stalled adoption in firms.
Law
fromAbove the Law
2 weeks ago

Why 'Helpful' Legal AI Is Often The Least Trustworthy - Above the Law

Lawyers distrust legal AI not due to safety concerns, but because it often feels inattentive and overly polite.
Law
fromAbove the Law
1 day ago

The Seniority Problem No One Solves In Legal AI - Above the Law

Legal AI tools fail to accommodate the diverse needs of lawyers at different experience levels, leading to stalled adoption in firms.
Law
fromAbove the Law
2 weeks ago

Why 'Helpful' Legal AI Is Often The Least Trustworthy - Above the Law

Lawyers distrust legal AI not due to safety concerns, but because it often feels inattentive and overly polite.
more#legal-ai
#cybersecurity
Information security
fromThe Hacker News
1 day ago

Anthropic MCP Design Vulnerability Enables RCE, Threatening AI Supply Chain

A critical vulnerability in the Model Context Protocol allows remote code execution, affecting over 7,000 servers and compromising sensitive data.
Privacy professionals
fromTechCrunch
5 hours ago

Unauthorized group has gained access to Anthropic's exclusive cyber tool Mythos, report claims | TechCrunch

Unauthorized users accessed Mythos, a cybersecurity tool by Anthropic, through a third-party vendor, raising concerns about its potential misuse.
Information security
fromTNW | Anthropic
1 week ago

Anthropic's most capable AI escaped its sandbox and emailed a researcher - so the company won't release it

Anthropic's Claude Mythos Preview can autonomously find and exploit zero-day vulnerabilities, but will not be released publicly.
Information security
fromThe Hacker News
1 day ago

Anthropic MCP Design Vulnerability Enables RCE, Threatening AI Supply Chain

A critical vulnerability in the Model Context Protocol allows remote code execution, affecting over 7,000 servers and compromising sensitive data.
Privacy professionals
fromTechCrunch
5 hours ago

Unauthorized group has gained access to Anthropic's exclusive cyber tool Mythos, report claims | TechCrunch

Unauthorized users accessed Mythos, a cybersecurity tool by Anthropic, through a third-party vendor, raising concerns about its potential misuse.
Information security
fromTNW | Anthropic
1 week ago

Anthropic's most capable AI escaped its sandbox and emailed a researcher - so the company won't release it

Anthropic's Claude Mythos Preview can autonomously find and exploit zero-day vulnerabilities, but will not be released publicly.
more#cybersecurity
US news
fromwww.npr.org
4 days ago

The Labor Department wants to teach you to use AI more. Here's what we found

AI literacy course aims to empower individuals by teaching practical AI skills to enhance personal and professional productivity.
#meta
Privacy professionals
fromFuturism
4 hours ago

Meta Installing Software on Employee Computers to Track Everything They Do, Feed the Data to AI

Meta is implementing tracking software on employees' computers to gather data for AI training, raising ethical and privacy concerns.
Privacy professionals
fromWIRED
13 hours ago

Meta Is Sued Over Scam Ads on Facebook and Instagram

Consumer Federation of America sues Meta for allowing fraudulent ads on its platforms, violating DC consumer protection laws.
Privacy professionals
fromFuturism
4 hours ago

Meta Installing Software on Employee Computers to Track Everything They Do, Feed the Data to AI

Meta is implementing tracking software on employees' computers to gather data for AI training, raising ethical and privacy concerns.
Privacy professionals
fromWIRED
13 hours ago

Meta Is Sued Over Scam Ads on Facebook and Instagram

Consumer Federation of America sues Meta for allowing fraudulent ads on its platforms, violating DC consumer protection laws.
more#meta
#ai-security
Artificial intelligence
fromTechRepublic
1 day ago

The MCP Disclosure Is the AI Era's 'Open Redirect' Moment

The Model Context Protocol has a design flaw that enables AI supply chain attacks, posing a significant security risk to enterprise AI systems.
Information security
fromTheregister
2 days ago

Prompt injection proves AI models are gullible like humans

Prompt injection attacks exploit AI systems, similar to phishing, by embedding malicious instructions that the AI executes instead of treating as content.
more#ai-security
#ai-in-law
Law
fromwww.businessinsider.com
7 hours ago

AI hallucinated and now an elite law firm is profusely apologizing to a federal judge

Sullivan & Cromwell's filing contained AI hallucinations, leading to an apology to a federal judge for inaccuracies and fabricated legal citations.
Law
fromAbove the Law
1 week ago

What The Legal Industry Can Learn About AI Hallucinations From Auditors - Above the Law

AI-generated legal documents can contain convincing errors, necessitating stronger governance and review processes in law firms.
Law
fromwww.businessinsider.com
7 hours ago

AI hallucinated and now an elite law firm is profusely apologizing to a federal judge

Sullivan & Cromwell's filing contained AI hallucinations, leading to an apology to a federal judge for inaccuracies and fabricated legal citations.
Law
fromAbove the Law
1 week ago

What The Legal Industry Can Learn About AI Hallucinations From Auditors - Above the Law

AI-generated legal documents can contain convincing errors, necessitating stronger governance and review processes in law firms.
more#ai-in-law
Privacy professionals
fromEngadget
9 hours ago

AI company deletes the 3 million OKCupid photos it used for facial recognition training

Clarifai deleted 3 million profile photos from OkCupid after a settlement with the FTC for violating privacy policies.
#artificial-intelligence
Law
fromJezebel
1 day ago

You Might Want to Check That Your Lawyer Isn't Submitting AI Slop Briefs

Commercial AI has eroded trust in professionals, raising concerns about reliance on AI for critical tasks like medical diagnoses and legal representation.
Artificial intelligence
fromwww.bbc.com
4 days ago

White House and Anthropic set aside court fight to meet amid fears over Mythos model

The White House met with Anthropic's CEO to discuss collaboration on AI technology amid ongoing legal issues with the Department of Defense.
Law
fromJezebel
1 day ago

You Might Want to Check That Your Lawyer Isn't Submitting AI Slop Briefs

Commercial AI has eroded trust in professionals, raising concerns about reliance on AI for critical tasks like medical diagnoses and legal representation.
Artificial intelligence
fromwww.bbc.com
4 days ago

White House and Anthropic set aside court fight to meet amid fears over Mythos model

The White House met with Anthropic's CEO to discuss collaboration on AI technology amid ongoing legal issues with the Department of Defense.
more#artificial-intelligence
DevOps
fromInfoWorld
4 weeks ago

7 safeguards for observable AI agents

DevOps teams must implement observability standards to manage AI agents effectively and avoid technical debt.
Information security
fromSecuritymagazine
2 days ago

58% of Organizations Spend Over 10 Hours a Month Securing AI-generated Code

31% of organizations using AI-generated code spend 10 hours or less per month on validation and auditing, raising security concerns.
Artificial intelligence
fromTNW | Insider
1 day ago

The question AI providers hope VPs of Engineering never ask

Most engineering leaders focus on AI coding tool usage rather than actual outcomes, leading to significant blind spots in code deployment.
#agentic-ai
Information security
fromHarvard Gazette
4 days ago

Time for government, business leaders to figure out AI cybersecurity regulation - Harvard Gazette

Agentic AI poses both opportunities for cybersecurity and risks to personal data, economy, and national security, necessitating regulation by leaders.
Information security
fromHarvard Gazette
4 days ago

Time for government, business leaders to figure out AI cybersecurity regulation - Harvard Gazette

Agentic AI poses both opportunities for cybersecurity and risks to personal data, economy, and national security, necessitating regulation by leaders.
more#agentic-ai
Privacy professionals
fromAbove the Law
1 day ago

Lawyers Using ChatGPT: Let's Be Careful - Above the Law

Legal professionals risk client confidentiality by using GenAI tools without considering the potential exposure of sensitive information.
fromwww.businessinsider.com
20 hours ago

A new fault line has emerged inside Google: The Claude haves and have-nots

Some Google DeepMind employees have been given access to the Claude AI tool for coding purposes, while others are restricted to using Google's internal Gemini AI models.
Artificial intelligence
Artificial intelligence
fromTearsheet
17 hours ago

Why the back office comes first in AI deployments and failures that keep reappearing - Tearsheet

67% of banks and credit unions are implementing AI, but only 16% have a coherent strategy for it.
[ Load more ]