The HAL dilemma: Why AI obedience may be more dangerous than AI rebellion | Brookings
The real danger of artificial intelligence is not AI thinking for itself—it's AI doing exactly what we ask, says Eduardo Levy Yeyati.
Reporting on AI risk: safety, security, bias, copyright, geopolitics and governance.
The real danger of artificial intelligence is not AI thinking for itself—it's AI doing exactly what we ask, says Eduardo Levy Yeyati.
... AI chips, to challenge Nvidia's ... Screen for heightened risk individual and entities globally to help uncover hidden risks in business relationships ...
AI is increasing service firm productivity, but also amplifying risks . Fortunately AI can also assist with governance.
A winner of this year's Nobel prize in economics warned Monday that artificial intelligence offers "amazing possibilities" but should be regulated ...
The AI Act categorises AI systems by risk , with high- risk financial ... The AI Act requires high-quality datasets to minimise risks of ...
Regulation & PolicyGovernance & Risk ManagementBanking & Finance
... AI tools, including any limitations on, and risks of, their use. ... Overall, the Guidance's strategies can help health care organizations minimize AI ...
The analyst noted that large-scale AI infrastructure deployments carry inherent risks of overbuilding, as companies compete to protect existing market ...
“Tempo's LogLM turns raw telemetry into high-signal context, the insight security teams need to outpace agentic AI threats , zero-click exploits and ...
In this regard, existing laws likely already address many of the high- risk concerns, such as the use of AI in the financial services sector, as the ...
In April 2025, my colleague Russ Anderson and I published an article discussing the potential risks associated with human resources (HR) ...
AT&T CISO Rich Baich shares five questions cyber leaders should answer as they look to build or strengthen their AI security program.
How to apply effective governance to harness the benefits of A.I. and mitigate its risks ... AI risks and incidents · Access to justice · Administrative ...
This working paper examines the risks and opportunities of artificial intelligence ( AI ) in Latin America and the Caribbean (LAC), with a special focus ...
The artificial intelligence boom is entering a new and riskier phase. Global capital expenditure on chips, data centres and cloud infrastructure ...
Bias in AI systems can lead to unfair treatment. An autonomous hiring agent could unknowingly filter out qualified candidates based on biased data.
Over the past two years, more candidates have figured out that employers aren't the only ones using AI ... Bias audits and governance: document how ...
Bias & FairnessGovernance & Risk ManagementJobs & the Economy
IRCC's website notes that “identifying and mitigating potential bias ... AI may replicate historical biases , leading to unfair outcomes for certain ...
AI learns from data, and if that data is biased , the AI might continue to make the same biased decisions. Some systems could still be biased toward ...
... AI prevent algorithmic bias – meanwhile may set a standard that requires AI developers to develop DEI frameworks, Sacks said, claiming that this ...
Bias & FairnessRegulation & PolicyGovernance & Risk Management
... Failures and Emergency Landings. Share. 0. 0. Lufthansa. Airbus A340-300. Engine Failure ... artificial intelligence ( AI ). Palantir faces valuation ...
KAIST launches nationwide campaign to share failures on International Day of Failure ... One new initiative, the AI × Failure Idea Contest, invites ...
The risks of bias ... Enhancing PUQ frameworks can help illuminate potential biases and ensure that AI models are equally reliable across diverse ...
Experts warn AI use in screening or negotiations can lead to liability. Human oversight and bias testing are critical for compliance. Legal risks ...
Bias & FairnessRegulation & PolicyGovernance & Risk ManagementLegal & Courts
These findings extend theories of cognitive offloading and automation bias by revealing the dual role of information literacy in AI -supported learning ...
Four models, GPT-4o, OpenAI o3, and the newer GPT-5 pair were put through the wringer. Then, another AI model graded their answers using a rubric that ...
In August, as part of the settlement, Starbuck was appointed to advise Meta on how to prevent political bias in its AI , The Wall Street Journal ...
IIT-Madras developed IndiCASA, a dataset to evaluate AI biases in India, focusing on caste, gender, and religion, addressing gaps in existing ...
You get a really super- biased AI .” A study, titled “Gender Bias in LLMs for Long-Term Care,” analyzed large language models – or AI systems that not ...
The 2025 Economics Nobel highlights how innovation drives growth and warns of risks from AI concentration, closed markets, and climate limits if ...
Our gravest AI risk isn't runaway superintelligence but our cultural readiness to surrender collective judgment to algorithmic systems.