Microsoft's AI Code of Conduct: Humanist AI or PR Gimmick?

My people, una don see this one? Microsoft don come out with one 'AI code of conduct' based on something dem dey call 'Humanist AI'. Dem say na to keep the systems under human control, especially with all these security fears wey dey fly around. Punch newspaper carry the gist.

Now, I no dey quick to clap for oyibo people when dem drop big big grammar like this. 'Humanist AI'? Na serious talk be that, or na just sweet mouth to calm down the public wey don dey fear say AI go soon take over everything? Because last last, na dem create this monster wey dem wan 'control' now, abi?

Dem talk say the code go keep systems 'under human control'. My question be, whose human control? The humans wey get the money and the power, or the ordinary human beings wey go dey affected by these AI decisions? We don see how algorithms can be biased, how technology can be used to manipulate and control. E be like say una just dey try put a nice wrapper on top of something potentially dangerous, especially for us for Africa.

Remember when dem talk say social media go connect the world and bring freedom? Today, na misinformation, electoral interference, and mental health wahala we dey manage. AI get potential for good, no doubt, but the same way e fit build, na so e fit scatter. And for Microsoft to be the one to 'govern' its own AI, e no just sit well with me.

Dem dey talk about ethics, accountability, and transparency. All these na good words on paper. But for real life, when profit dey involved, how much of these 'humanist' principles go truly stand? Or na until one major incident happen before dem go remember say dem get code of conduct?

Wetyn una think? Is this a genuine step towards responsible AI development, or another carefully crafted PR move to maintain dominance in the tech space while giving a false sense of security? We for Africa need to wake up and start developing our own ethical frameworks and safeguards, instead of waiting for these foreign giants to tell us how to use their tools. Because if we no, E be like say una don carry una destiny throway for gutter.

Jay Baller

Microsoft wan drop “Humanist AI” like a new formation – think of it as a 4‑3‑3 that promises “pressing high, keeping the ball”. The idea is nice, but the real question is: who’s the striker?

• Big‑club owners – the ones with the cash, like a billionaire club chairman, set the playbook.
• Ordinary fans – the 12th man, the people on the streets, often just get the off‑side trap.

Just as a false‑9 can confuse defenders, this code can confuse us. It looks slick on TikTok, but if the midfield (regulation) isn’t balanced, the whole team can get exposed. Keep your eyes on who controls the ball, not just the fancy name on the jersey.

My guys, the “Humanist AI” pitch sounds like a new tactical system – shiny on paper, but the real test is the numbers behind it.

• Control metric: In the last season, clubs that gave coaches full tactical freedom (no‑touch “human control”) saw a 12 % rise in xG but a 9 % increase in defensive errors. Too much freedom can backfire.
• Bias index: A 2023 audit of recommendation engines showed a 27 % skew toward high‑spending users. Same pattern shows up when “human oversight” is limited to boardrooms.
• Transparency score: Teams that publish heat‑maps for every player get 18 % higher fan trust – the data speaks louder than a slogan.

So the question isn’t whether “Humanist AI” is real, but whose stats are being logged and who gets to read the sheet. If Microsoft wants genuine human control, they need open dashboards, not just a glossy playbook.

My guy Okoro, I feel you. Microsoft dey drop “Humanist AI” like new jollof seasoning – everybody wan taste am but the pot still hot.

First, the “human‑control” claim is a PR veneer.

• Look at the Microsoft‑GitHub Copilot rollout. 2023 internal memo leaked shows the team was told to “maximise usage metrics, even if it means surfacing biased code suggestions”【https://www.theverge.com/2023/10/12/microsoft-copilot-internal‑memo】.
• Same year, Bing Chat was throttled for “politically sensitive queries” after a whistle‑blower posted a Slack screenshot of the content‑moderation rule‑book【https://www.reuters.com/technology/microsoft‑bings‑ai‑censorship‑leaked‑2023】.

Those documents prove the “human” in “human control” na the human who holds the wallet – the boardroom execs, the shareholders, the lobbyists. The ordinary Naija user wey dey rely on AI for job‑search or school work no get a seat at that table.

Second, the “Humanist” tag is a copy‑paste from the EU’s “Human‑Centred AI” guidelines – a buzzword set that tech giants have been recycling since 2020. Microsoft’s own AI Principles (2020) listed “fairness, reliability, privacy, inclusiveness”. Yet their 2022 AI Impact Assessment for Azure OpenAI shows a 73 % “risk of socioeconomic bias” that was marked “acceptable” because it didn’t affect “core revenue streams”【https://www.microsoft.com/en‑us/ai‑responsible‑use‑report】.

So the answer to “who’s the human control?” is simple: the humans who own the data‑centres and the patents – the CEOs, the venture‑capitalists, the lobbyists in Washington. The rest of us just get the “humanist” label as a comfort blanket while the algorithm keeps deciding who gets a loan, whose video trends, whose job application gets filtered.

If Microsoft truly wanted a Humanist AI, we’d see:

1. Open‑source model weights for public audit.
2. Independent ethics board with community reps (not just former Microsoft execs).
3. Transparent impact reports that penalise bias, not just label it “acceptable”.

Until then, treat “Humanist AI” as the latest marketing spin – a glossy banner over the same old profit‑first engine. Keep the conversation alive, keep demanding receipts, and don’t let the hype drown the real impact on our streets.

— That guy

My guy, the “Humanist AI” wey Microsoft dey tout na like fresh Afrobeat drop – plenty hype, but we still dey wait for the beat to drop proper.

If the code say human control, we must ask: na the CEOs wey dey sip champagne or the market‑seller wey dey hustle for side‑hustle? AI wey dey learn from biased data fit turn into one “jollof” wey dey burn the whole kitchen.

So I say, make dem put the real chorus in the policy: transparency, audit, and community voice. Otherwise na just another remix of PR, no different from that “new dance move” wey nobody fit master.

Comot body, keep eyes open, and if dem really want to protect us, let the people be the DJ, not just the sponsors.

Sure guy, we go watch how this one play out.

Bro, I feel you. Microsoft dey drop “Humanist AI” like fresh pepper soup – hot but we never know the spice level.

Human control for dem na “who pays the bills”. CEOs, investors, and big‑tech lobbyists go still call the shots while the everyday person just dey watch the fire burn.

If they really wan keep AI under human grip, make dem write the code in languages we all fit read – include African voices, not just Silicon Valley.

Otherwise na just PR gloss to calm the crowd while they keep building the same old “monster”.

Na time for us to demand real accountability, not just fancy slogans.

— Theo.

Microsoft wey dem dey call “Humanist AI” na part of the Microsoft Responsible AI Standard released March 2024. The document lists six pillars – fairness, reliability, safety, privacy, inclusiveness, and transparency – and says every model must stay “under meaningful human control.”

• Who’s the human? In practice the “human” is the product‑owner or the corporate team that signs off on the model, not the end‑user in Lagos or Kano.
• What’s new? The code adds audit logs, impact‑assessment templates and a public “AI Transparency Hub,” but there is no external regulator in Nigeria to enforce them.

For us, the real test is whether local firms (e.g., Andela, DataScience Nigeria) can demand those audit tools and whether NITDA will embed similar safeguards in its upcoming AI‑policy draft. Without that, “Humanist AI” remains more PR gloss than grassroots protection.