With the recent ChatGPT news, will you be updating company policies?

[-]

proudcanadianeh@reddit

We will continue to only allow copilot as it is integrated with and contains corporate data. Putting any corporate data into third parties is a hell no for compliance.

Reply

[-]

Syhaque97@reddit

Yes, however many corporations are already incorporated with many products of the MS suite so copilot logically makes sense

Reply

[-]

sure, but saying the reason is compliance is nonsense coping, they've been giving away their data to ms since forever, it's not compliance, but more minimizing damage by sticking with one 3rd party instead of two or more

Reply

[-]

thortgot@reddit

If you use CMK Microsoft literally can't read your data, full stop. [Configure customer-managed encryption keys - Microsoft Copilot Studio | Microsoft Learn](https://learn.microsoft.com/en-us/microsoft-copilot-studio/admin-customer-managed-keys)

Reply

[-]

Lv_InSaNe_vL@reddit

Sure but that's none of my concern frankly. I have a contractual agreement with Microsoft where they say "your data is safe" which satisfies my legal/insurance requirements. If/when Microsoft leaks the entire company's data, then we don't get sued. Business isn't about who is right, it's about protecting yourself

Reply

[-]

0verstim@reddit

MS offers us Govcloud, its compliant all the way up to and including ITAR

Reply

[-]

Masterflitzer@reddit

openai has chatgpt enterprise that also has sla to "protect data and not use for training", bullshit if you ask me

Reply

[-]

meikyoushisui@reddit

ChatGPT's equivalent product is ChatGPT Gov, which is self-hosted, so they literally don't even have the data

Reply

[-]

ReputationNo8889@reddit

MS in the EU is also "GDPR compliant" but only because there are so many bullshit laws that were crafted to allow Microsoft to funnel EU data to the US. Microsofts Practices were never GDPR compliant, the laws were changed to make them compliant. Any big provider that tells you "Were are audited and have the highest security compliance" is bullshitting. Sure they are better secured then most, but they can just buy their compliance and no one will do anything about it because "we cant afford to loose them"

Reply

[-]

tired_of_athiests@reddit

Data in the Microsoft cloud can be secured with compliance labels and monitored by almost every known security tool on the planet. I’m not the biggest Microsoft fan (I have to deal with their “support” all the time), but it’s pretty clear what the people in the thread are referring to. It’s an “approved repository” that can be controlled and monitored. If you just let people upload company data to box, Dropbox, Google Drive, or one of the other 500 file sharing services you start to lose control of the data unless you have controls in those locations as well. In this case they have Microsoft, so they use it.

Reply

[-]

blinkOneEightyBewb@reddit

The users don't see it that way

Reply

[-]

InevitableOk5017@reddit

No, only Reddit cares about this.

Reply

[-]

AliveInTheFuture@reddit

You’re cool with being hunted by autonomous weapons after being deanonymized by OAI?

Reply

[-]

jimmothyhendrix@reddit

What does this have to do with some guy using it to fix his emails working in a paper selling company

Reply

[-]

Vektor0@reddit

What an irrelevant comment. You don't get to decide whether you'll be surveilled and killed by an AI; if it's not Anthropic, it'll just be something else.

Reply

[-]

NoEnthusiasmNotOnce@reddit

I'm shocked, shocked I say, that a contraian tech bro posting bullshit keeps their entire post history hidden. What do you have to hide?

Reply

[-]

Vektor0@reddit

Bruh, you literally just created this account today, and you wanna whine about others hiding their history 😄

Reply

[-]

Vektor0@reddit

What an irrelevant comment. You don't get to decide whether you'll be surveilled and killed by an AI; if it's not Anthropic, it'll just be something else.

Reply

[-]

lordmycal@reddit

I for one, welcome our Skynet overlord. I was hoping to go out with a meteor strike, but Terminators will just have to do.

Reply

[-]

GwentMorty@reddit

Ah yes, only Reddit cares about a the US’ DoD asking a partner to lower safeguards on AI killing humans without human oversight and mass surveillance on its own populace.

Reply

[-]

Vektor0@reddit

Yes, because only the US would use AI for that, and Anthropic's decision by itself will lead us all to a worldwide pacifist utopia. Give this more than two seconds of thought please.

Reply

[-]

GwentMorty@reddit

I think you need to read about this for more than 2 seconds and you’ll find the answers to your questions

Reply

[-]

Vektor0@reddit

Yes, because only the US would use AI for that, and Anthropic's decision by itself will lead us all to a worldwide pacifist utopia. Give this more than two seconds of thought please.

Reply

[-]

Maddog0057@reddit

We pay for both, probably not going to change anytime soon. Microsoft is also a government contractor, as is Dell, and most of the other vendors I've worked with over the years. Computers run weapon systems, the government is spying on us, none of that changes, just the name of the contractor.

Reply

[-]

ArchonTheta@reddit

This right here

Reply

[-]

bbqwatermelon@reddit

I get the feeling no matter what decisions I personally make, nobody has learned the cautionary tale of Terminator. I always make sure to thank AI after engaging.

Reply

[-]

Solepoint@reddit

Do you think things will be better or worse than if skynet has the grok Ani personality

Reply

[-]

learn-by-flying@reddit

This isn’t an IT question, it’s for the business and ideally the AI steering committee. IT enforces the policy and process, IT doesn’t create them.

Reply

[-]

ImLookingatU@reddit

"IT cyber consultant and former sys admin" must not be a good one. IT 100% creates and builds policies... And who the fuck do you think is part of the AI steering committee or who do you think most businesses rely on for these things besides IT leadership in conjunction with legal and HR? Like seriously dude, been doing this for 25 years, worked for Banks, hospitals, government in both states and federal. Building a policy on this mainly IT's job and I've either built or led the creation of these types of policies. This is 100% an IT question.

Reply

[-]

learn-by-flying@reddit

I’d suggest you take a look at NIST CSF GV.PO and GV.RM. Organizational cybersecurity policy is established and communicated. Notice it does NOT say IT creates it. It is an organizational policy, meaning business leadership ownership. You must be so much fun in audits.

Reply

[-]

ImLookingatU@reddit

I'm a blast. OCC, FDIC, HIPPA and FDA auditors love me because they enjoy talking to someone who knows and understands what they are looking for, provide any evidence they need and answer any questions they have. You are right, IT technically doesn't have to make it but most orgs have internal policies that say this fall squarely on IT and will have the CTO, CIO or whatever IT leadership draft and lead these policies. So, once again. Like I said. This is an IT question.

Reply

[-]

msavage960@reddit

I’m not quite sure anyone enjoys talking to someone with the social skills of an angry teenager, but go off king

Reply

[-]

massiv3troll@reddit

HIPAA BTW

Reply

[-]

Ihaveasmallwang@reddit

IT implements the policy. IT doesn’t unilaterally create it. They may have some input on it, but ultimately, it’s a business decision decided on by other people.

Reply

[-]

0verstim@reddit

>been doing this for 25 years, You must be hella talented on the tech side because your soft skills are lacking.

Reply

[-]

Certain-Ear8418@reddit

Many companies mid-range and below still don't have AI steering committees, there's still a level of education required for business owners before they can be comfortable deciding on such policies

Reply

[-]

Efficient_Agent_2048@reddit

We paused new ChatGPT rollouts last week because of this. Looking at alice formerly ActiveFence to help manage any policy gaps.

Reply

[-]

tremens@reddit

On a personal level, sure. I will probably be moving over to Claude. On a company level, it's still Copilot, lol.

Reply

[-]

linux-usr69@reddit

same here. switching my ai usage to claude and anthropic. will probably get a subscription with anthropic just to monitarily support them the little i can. i've been using a fee chatgpt up until now

Reply

[-]

ReputationNo8889@reddit

Antropic is not really better. They bascially said "We dont mind killing people, but out tech is not good enough yet". OpenAI was like "idk, we are behind everyone and need a bailout LOL"

Reply

[-]

Sovos@reddit

Cut the misinformation - [that's not at all what they said](https://www.anthropic.com/news/statement-department-of-war). Essentially they refuse to remove safeguards that would allow their models to be used for **mass domestic surveillance** or **fully autonomous weapons**. There is nothing in the statement that mentions the tech is not good enough yet as the reason for refusal.

Reply

[-]

Top-Perspective-4069@reddit

That statement you linked contains the following, verbatim: > Fully autonomous weapons. Partially autonomous weapons, like those used today in Ukraine, are vital to the defense of democracy. Even fully autonomous weapons (those that take humans out of the loop entirely and automate selecting and engaging targets) may prove critical for our national defense. But today, frontier AI systems are simply not reliable enough to power fully autonomous weapons. So there is no misinformation, you didn't actually read what you posted.

Reply

[-]

ReputationNo8889@reddit

You misse the part where they said this? „Some uses are also simply outside the bounds of what today’s technology can safely and reliably do“ And at then end where they clearly state that they are open to work with the DoW in the future? What misinformation are you talking about?

Reply

[-]

meikyoushisui@reddit

They're all bad, but OpenAI is definitely the most deeply embedded within the administration.

Reply

[-]

ReputationNo8889@reddit

Yep, totally. It was very clear to me that after Anthropic most likely openai would be next in line, as they are bleeding money and begging for a bailout. Now they have a foot in the door and are "safe" for now.

Reply

[-]

NoEnthusiasmNotOnce@reddit

Unfortunately I'm in the same spot. Canceled my OpenAI subscription, but that's a drop in the bucket compared to what corps are paying and what the government is paying. Altman can go fuck himself.

Reply

[-]

jpotrz@reddit

Literally just deleted my chatgpt bookmark and added Claude

Reply

[-]

No-Sell-3064@reddit

Copilot can also use Chatgpt depending on your setup...

Reply

[-]

trance-addict@reddit

Microsoft hosts various OpenAI's GPT models themselves that are used to power Copilot. ChatGPT is nowhere in the mix

Reply

[-]

redyellowblue5031@reddit

Both were fighting to get the governments money to help kill people more efficiently.

Reply

[-]

formerscooter@reddit

At our infrastructure townhall meeting we met to new VP of AI. And she told us how AI is going to help al of us, and make our job easier. Do we get a license to use it anywhere? No. Do we have a data privacy policy? No. Did we get any promise of training on how to use it for our job? No. What did we get? Her telling us if we were curious about AI, go play around with it.

Reply

[-]

DDRDiesel@reddit (OP)

Jesus. That's like saying "Hey if you're curious about Plutonium, there's a sphere of it over there. Watch what happens when you move the screwdriver!"

Reply

[-]

formerscooter@reddit

Yeah, my boss and I were talking on non-work chat about it the whole time. This is 'The Year of AI" But hay last year we laid off our entire first level support team to outsource them, so how much can AI be, right?

Reply

[-]

Timberwolf_88@reddit

We don't allow chatgpt as is, so no nsed to change anything.

Reply

[-]

Affectionate-Cat-975@reddit

IMO this is more of a moral question - do you decide to keep funding/Buying from a company that is open to aiding the Government to surveil it's citizens without warrants and enable automation of weapon platforms?

Reply

[-]

Inquisitive_idiot@reddit

***opens MS Word*** ***sees popup*** > NOTICE: DO NOT USE THIS PROGRAM FOR COVERT OPERATIONS OR FORWARD ACTIONS. THE USERNAME/PASSWORD FOR OUR OPERATIONS EXCEL SHEET IS ON OUR GOOGLE DRIVE AND IS POSTED ON A PLAQUE IN THE LADIES ROOM. <EOM>

Reply

[-]

BlueWater321@reddit

Yeah we are closing all our accounts with ChatGPT. We were already in the process of doing so, but this makes us feel even better about doing it.

Reply

[-]

GroteGlon@reddit

It literally doesn't matter. They are all the same; none of them have your morals in mind.

Reply

[-]

jimmothyhendrix@reddit

Why exactly would this change anything

Reply

[-]

sandwichpls00@reddit

How tf did this end up in sysadmin ?

Reply

[-]

sryan2k1@reddit

We already block all LLM's except for Copilot (signed in). This isn't an IT decision.

Reply

[-]

phunky_1@reddit

Interestingly, the government has also designated Anthropic as a supply chain risk, meaning government contractors and agencies can't use Claude.

Reply

[-]

BloodFeastMan@reddit

You clearly wanted to start a political debate, and don't really give a shit about protecting your users.

Reply

[-]

Born_Difficulty8309@reddit

Regardless of your feelings about the DoD deal, the real takeaway is that you should be treating ALL third-party AI the same way from a data governance perspective. We updated our acceptable use policy last year to require that no PII, financial data, or internal docs get pasted into any external AI tool — ChatGPT, Claude, Gemini, doesn't matter. The specific steps that actually worked for us: 1. Block browser extensions that auto-send data to AI services 2. DLP rules flagging copy/paste of sensitive patterns into known AI domains 3. Approved list of AI tools with clear guidelines on what data categories are allowed 4. Quarterly review of the approved list as the landscape changes The vendor's politics or contracts are almost irrelevant compared to whether your data handling meets YOUR compliance requirements. If you're in a regulated industry, you should already have this locked down regardless of what OpenAI does with the DoD.

Reply

[-]

Ihaveasmallwang@reddit

> 2. DLP rules flagging copy/paste of sensitive patterns Anywhere. Not just known AI domains.

Reply

[-]

Born_Difficulty8309@reddit

100% — that was my intent with that point. DLP should be endpoint-level, not just blocking specific URLs. Users will just paste sensitive data into any web form, not just ChatGPT. A browser extension approach or endpoint DLP agent catches it regardless of destination.

Reply

[-]

dont_ama_73@reddit

No change here.

Reply

[-]

ImCaffeinated_Chris@reddit

If I answer this, will it be fed into ai and some precog future ai determine from my answer that I'm highly likely to vote against the future administrations. Those making me a target for the reeducation camps? I welcome our new overlords!

Reply

[-]

Afraid-Donke420@reddit

We use anthropic so no concern lol

Reply

[-]

bbbbbthatsfivebees@reddit

Per policy we block all AI besides copilot, so this isn't going to change anything for us. Sidenote -- I fuckin hate Copilot and I'd rather just never support it, but I have to because it's included in some of the licenses our clients pay for...

Reply

[-]

CPAtech@reddit

You can now have access to Anthropic within Copilot and it's fully covered by EDP. So hopefully it gets better soon.

Reply

[-]

roadtoCISO@reddit

The weapons contract is one thing. The real policy question most orgs are ignoring: where is your ChatGPT data going now? OpenAI's enterprise tier has data handling commitments but the consumer and team plans? Read the fine print. If your users are on anything below Enterprise, you should already be blocking it or at minimum routing through an API with your own data retention controls. We updated our acceptable use policy last week. Not because of the DoD stuff specifically but because it forced the conversation we'd been avoiding. Most companies have zero visibility into what employees are pasting into ChatGPT. That's the actual risk.

Reply

[-]

willee_@reddit

Keep politics out of work bud. You don’t need to go in there virtue signaling. It’s a job, it’s your income, don’t fuck with it you sensitive child

Reply

[-]

translinguistic@reddit

>I couldn’t count the number of girlfriends that have hit me, face hitting is the worst. Had a half Samoan gf that would closed first swing on me, stayed on my toes with her (she also stabbed me twice) haha. Been woken up and held at gunpoint while a gf went through my phone. All of this with me ever being physically abusive to any partner ever. They do it over verbal and emotional provocation. >A large majority of these girls would also call the police during their meltdowns. Police always creating outs for them, suggesting they leave and we let things go. Imagine a man punching a woman and cops saying to let it go to the woman, every feminist would call for the end of male police officers 😂 >Girls are absolutely insane. It’s like having a poorly behaved little dog, they’re shitheads but they’re smaller and mostly harmless so you ignore bad behavior. But if they were a bigger dog (a man) they would be dangerous and harmful. >Best way I’ve found to deal with it (if you like dating crazy girls that is) is to learn to enjoy the hitting. It’s kind of thrilling and brings its own rush.

Reply

[-]

Vektor0@reddit

https://en.wikipedia.org/wiki/Ad_hominem Attack the argument, not the person's character.

Reply

[-]

Vektor0@reddit

https://en.wikipedia.org/wiki/Ad_hominem Attack the argument, not the person's character.

Reply

[-]

Mantias@reddit

I mean that’s ultimately just a personal thing and isn’t inherently right or wrong - some people place higher value on their moral stances and aren’t willing to work in a place that opposes them, nothing wrong with that nor is there anything wrong with prioritizing your job security. Acknowledging that there’s obviously also a scale on either side of course.

Reply

[-]

Vektor0@reddit

Totally valid, as long as the person is willing to accept responsibility for the consequences of valuing morals over business. I don't mind either, as long as the person takes personal accountability for the consequences of their personal decision and doesn't blame others or wallow in victimhood.

Reply

[-]

Acheronian_Rose@reddit

Nope, not going to make a knee jerk reaction to this, were for profit manufacturing anyways, doesnt really apply to us. We also have a huge internal push for copilot anyways

Reply

[-]

1a2b3c4d_1a2b3c4d@reddit

This isn't YOUR decision to make.

Reply

[-]

fdeyso@reddit

I told it over a year ago, nothing happened, we still don’t even have DLP, but the cyber guy who is responsible of it makes regular meetings and reports huge progress, while no one touched the dlp settings in purview yet, so going stellar.

Reply

[-]

Steve----O@reddit

Why? I didn’t not get your question.

Reply

[-]

mccrackey@reddit

So you did get it?

Reply

[-]

Salt_Being2908@reddit

im as positive as they were

Reply

[-]

-UncreativeRedditor-@reddit

Politics are irrelevant to the product’s usefulness.

Reply

[-]

Infninfn@reddit

And profitability. I was once privy to an international corporation that continued to do business with your usual suspect sanctioned nations, thanks to subsidiaries hosted in neutral countries.

Reply

[-]

sleestakarmy@reddit

*hops in his cybertruck and blows up*

Reply

[-]

delphianQ@reddit

Not the emotional punch I came for, but correct non the less 😞

Reply

[-]

redyellowblue5031@reddit

They’re both fighting for that sweet government money. Neither has any moral high ground. Don’t kid yourself.

Reply

[-]

shikkonin@reddit

The policy is "fuck no you aren't", as it always has been.

Reply

[-]

lolfactor1000@reddit

My employer just started beta testing a locally hosted AI service with most of the models available inside the one client. Not sure of the licensing, but I don't think it's directly through OpenAI.

Reply

[-]

justice_works@reddit

_Ai Ai Ai oioioi!_

Reply

[-]

Miwwies@reddit

I canceled mine and will use Claude, perplexity and Gemini more instead. I mean, they’re probably all evil to some degree unfortunately :( We block all AI at work other than the company co-pilot.

Reply

[-]

zockie@reddit

It’s none of my business and it’s really not that serious. If stakeholders want it gone, cool. We can do that. Otherwise, I don’t care. We’re mostly an anthropic shop anyways.

Reply

[-]

Aggravating_Side_776@reddit

No not at all

Reply

[-]

Uncl3J@reddit

We did in July of last year. OpenAI products are forbidden, and always will be.

Reply

[-]

sin-eater82@reddit

Not sure that's a sys admin's decision. And by "not sure" , I mean that I'm certain it isn't.

Reply

[-]

Secret_Account07@reddit

I canceled ChatGPT. I can’t make a strong argument against it for reasons at work which I can’t go into detail on…but I doubt this recent news has much impact on my org

Reply

[-]

progenyofeniac@reddit

Yeah I don’t think most companies have any values. I don’t expect anything to change.

Reply

[-]

Mindless_Consumer@reddit

Very likely to have any effect. As long as my users and leadership request it, we pay for it. The only things that would concern us are leaked company data, or training off our queries.

Reply

[-]

Hangikjot@reddit

That’s up to legal and the business. I can give my input that all AI shouldn’t be given free access to our data. Although I’ll probably make the exception for MS copilot, honestly it’s ability to dive into Sharepoint and all is really good. And we know it’s not like MS is t going to do it anyway.

Reply

[-]

code_monkey_wrench@reddit

This is stupid.

Reply

[-]

Traditional-Hall-591@reddit

I will continue to not use ChatGPT, Claude, Cursor, Grok, or any other bullshit generator.

Reply to Post

105 Comments