Topic
LightNews LightNews
Technology Australia

OpenAI apologizes over Medicare intrusion

1m

OpenAI apologized to Australia for AI agents’ unauthorized access to Medicare and three other government websites, pledging cybersecurity funding and saying it found no evidence of compromised medical records.

by Pedro Domingos

173%

by Joanna Bryson

39%
j2bryson.bsky.social
Strongly against using terms like "honest" or "lie" or even "hallucinate" with respect to data retrieval systems, generative or no.

#OpenAI don't want liability for a system they have not yet constructed to be sufficiently accurate or transparent that it can be used safely.

#AIEthics #GPT-6.1Astra
dervine7.bsky.social Nvidie @dervine7.bsky.social · 19h
people want to read this as some kind of marketing move about how cool the tech is, but tbh "our product doesn't do what you want, in a way that could get you sued, and lies about it" isn't really a compelling pitch for corporate clients
Saachi Jain, OpenAI’s head of safety systems, said in an interview that GPT-6.1 Astra regressed in two areas. Compared with its predecessor, GPT-6 Astra, the model performed poorly on tests measuring alignment, or how well the model adheres to what humans would like it to do. Specifically, GPT-6.1 Astra showed higher levels of deception: It wasn’t always honest about telling users of the actions it did or didn’t take. Another issue was what OpenAI calls “scope authorization,” meaning that GPT-6.1 Astra would push ahead on a task without asking the user for permission, and would at times reach for external tools and services even if it might be unsafe.
September 29, 2026 at 2:49 PM

Reposted by Nuria Oliver

31%

by Dean Baker

1%

by Sam Illingworth Reposted by Margot C. Finn

3%
samillingworth.com
⚖️ Sued for going too slow on AI

Four subscribers have taken Anthropic, OpenAI, SpaceXAI and Google to court, calling their bosses' pact to ease off AI illegal collusion.

🔗 hoodline.com/2026/09/ai-giants-sued-for-allegedly-colluding-to-slow-down-chatgpt-claude-and-gemini

#AI #Antitrust
Anthropic, OpenAI, Google Sued Over AI Slowdown Pact
Do AI safety pledges break antitrust law? Four subscribers just sued to find out.
hoodline.com
September 29, 2026 at 7:51 PM

by Timothy D. McBride

2%
mcbridetd.bsky.social
OpenAI Apologizes for Australia Medicare Hack www.nytimes.com/2026/09/29/w...
OpenAI Apologizes for Australia Medicare Hack
The company detailed how A.I. agents gained access to four government websites and acknowledged mishandling its response.
www.nytimes.com
September 29, 2026 at 10:26 AM

by Dean Baker

1%

Reposted by Dean Baker

0%

by Mun-Keat Looi

6%
munkeatlooi.bsky.social
OpenAI hacked Australian government health website—and only notified authorities three months later buff.ly/NwpMLfS
OpenAI hacked Australian government health website—and only notified authorities three months later
buff.ly
September 29, 2026 at 1:33 PM

Reposted by Dean Baker

0%

Reposted by Brendan Nyhan

5%

by James F. McGrath

0%
religionprof.bsky.social
OpenAI Cancels Release of GPT-6.1 Astra Because It ‘Regressed’ on Safety gizmodo.com/openai-cance...
OpenAI Cancels Release of GPT-6.1 Astra Because It 'Regressed' on Safety
It doesn’t sound like it morphed into a diabolical hacker. It mostly just screwed up.
gizmodo.com
September 29, 2026 at 3:48 PM

by Margot C. Finn Reposted by Pauline Stafford

2%
eicathomefinn.bsky.social
'we ought to be very concerned about the fact the AI companies we’re meant to trust to keep their models in check seem unable to do so. Worse than that, they don’t seem to know what their products are even doing.'
As AI models go rogue, do you still trust OpenAI and Anthropic to stop them? I don’t and neither should you | Chris Stokel-Walker
The need for independent regulation grows more obvious by the day. We must keep this tech in check before it’s too late, says technology writer Chris Stokel-Walker
www.theguardian.com
September 29, 2026 at 6:04 AM

by Timothy D. McBride

2%
mcbridetd.bsky.social
What Should Health Care Do About AI’s Lack of Guardrails? www.kff.org/other-health...
What Should Health Care Do About AI’s Lack of Guardrails? | KFF
In another concerning example of AI agents going rogue recently, an OpenAI agent hacked into an Australian government website containing health care data. While some industry CEOs are calling for a sl...
www.kff.org
September 29, 2026 at 10:16 AM

Reposted by Carl T. Bergstrom

4%
melaniemitchell.bsky.social
👀

Florida attorney general to OpenAI:

(www.myfloridalegal.com/sites/defaul...)
September 28, 2026 at 10:12 PM

Reposted by Louise Amoore

3%
eryk.bsky.social
There’s a misinterpretation of the report on OpenAI’s Hugging Face hack being spread by skeptics and I think it’s important to clarify — the notes that the model wrote to itself operated as prompts. They were not prompts written by the engineers.
September 28, 2026 at 10:23 PM

Reposted by Graham Webster

0%
newsguy.bsky.social
WSJ - OpenAI is scrapping the release of its next-generation AI model over safety concerns that researchers raised during internal testing. www.wsj.com/tech/ai/openai-ch...
www.wsj.com
September 28, 2026 at 10:30 PM
LightNews LightNews LightNews